Re: [Pearg] Research Group Last Call for "A Survey of Worldwide Censorship Techniques"

[Niels ten Oever <[e-mail]>]

On 5/29/20 10:34 AM, Eliot Lear wrote:
> Hi Chris,
> 
> 
>> To try and sharpen this point a bit: this document will certainly have gaps, and it's not a goal to fill all those gaps. This field moves quickly enough that trying to doing so would likely be futile. 
> 
> I agree, and I am not suggesting that all gaps be covered.  That was a straw man put forth by the author.
> 
>> The chair decision to begin RGLC was based on the document's content and history, which was deemed sufficiently thorough with respect to relevant issues that protocol designers might consider going forward. That said, concrete suggestions to fill in those gaps are welcome! The authors and RG should certainly review them if provided, though we won't block progress waiting for additional content that doesn't substantially change the document's purpose.
> 
> I am sorry I didn’t catch this earlier, but I presumed that last call was meant to address these sorts of matters.  I would ask that it be noted that there is not a firm consensus on the document as it stands, as several of us have raised substantial issues.  Were I to explore them outside of this group in the broader IETF or research communities, I feel certain that you would not find consensus over the terminology.  You needn’t take my word for this.
> 
> To restate my concerns, there are three major points and one further down that may or may not be a major point:
> 
>   * The definition of censorship itself does not match common usage.

I disagree. Saying something doesn't match common usage, or common sense, it easy to say but very hard to show. To define what censorship is, is not what the authors set out to do, and I don't think we can reasonably ask it of them. 

>   * There is no definition of “censorship regime”.

I don't think that is needed because this is not a literature review on censorship. The study provides a definition and implements it. Very typical for a survey study. Else you will have a chicken and egg situation: should the definition depend on empirical evidence, but how to do case selection without a definition?

>   * The one reference I checked was clearly not appropriate in its usage, leading me to wonder about the others.

We should not depend on it, but this is also fully checked in the RFC editor process.

Best,

Niels

> 
> 
> On terminology
> 
> It seems that it is being claimed that these are terms of art.  Regardless, a properly justified definition with an appropriate reference is in order.  This has not happened here.  There are two specific problems with the first sentence and its reference.  First, the text doesn’t match the reference given.  It is never appropriate to mislead the reader in this way.  Second, Wikipedia is not a primary source.  It is especially important to get the definition right in interdisciplinary work, especially when we are talking about defining a very politically charged term that the rest of the document is dedicated to ameliorating.  Beyond the academic matter, as I have previously alluded, by not setting that context, the authors risk offending many practitioners dedicated to infrastructure protection in the very first sentence of the first section.
> 
> Unless you find fault with my reasoning above, and I readily allow for such a possibility, we must agree that this is an issue.  If you place the onus on me to solve it and I refuse, the issue would remain.  Of course I am happy to be a part of solving the problem.  As it happens, Joe himself put forth a perfectly fine way to put the work into context, to which I readily agreed, but then he himself rejected, even though what he proposed is a close variant of what appears in the document’s abstract.  The substance here is the difference between “content" and “information” and perhaps motivation.  I am not suggesting a full exploration of what censorship is, but some justification for the language.  Same for “censorship regime”, given its similar negative connotations.
> 
> On references
> 
> We have previously seen sloppy use of references in at least one other RG.  By chance, I have spotted two errors, one to the WP reference and the other only because I know one of the authors of the referenced work.  I have not checked the others, but the pass rate right now is 0/2.  I would strongly suggest that the other references be checked through broader review.
> 
> Adding the additional non-technical example
> 
> On “Non-Technical [Prescription/Interference]”.  As I wrote above, I don’t expect the authors to be exhaustive, but I am told that the topic is out of scope, and yet the text remains.  I would suggest an add, perhaps around “swaying public thought”, to include something like “adding financial liability”.  The reason I think the point is important enough to raise is that from a scale perspective, the financial attack may be the most effective attack of all against profit-based platforms.  Is it major or minor?  I don’t know.  I lean toward “minor” only because the document might stand up without its mention, even though I would view it as an opportunity lost.
> 
> Eliot
> 
> 

-- 
Niels ten Oever
Researcher and PhD Candidate
Datactive Research Group
University of Amsterdam

PGP fingerprint	   2458 0B70 5C4A FD8A 9488  
                   643A 0ED8 3F3A 468A C8B3