Arnaud’s Open blog

Opinions on open source and standards

What consensus means

Over the last year I’ve noticed that quite a few people are using the word “consensus” in a way which differs from my understanding of what consensus is about.

Looking at Wikipedia, I see that it defines consensus as “general agreement”. This is pretty vague obviously, and if it were left to that there wouldn’t be much more to talk about. But Wikipedia quite rightly points out that consensus is also used to refer to the process used to reach this agreement. I think that’s where the problem lies.

From what I can tell, some people are happy to just use consensus as if it were synonymous with agreement. On that basis, any decision, no matter how it is made, can pretty much be said to be the consensus.

For instance, in the case of the BRM for OOXML, it has been stated by several, and ISO/IEC officials in particular, that the decisions were made by “consensus”. Was it so, though? I certainly don’t think so.

I’ll admit that my expectations are heavily rooted in my background with the W3C which inherited from IETF the goal of making all its decisions by consensus. As I explained in my previous entry A Standards Quality Case Study: W3C, the W3C leaves it to the chair to decide whether consensus is reached or not.

Interestingly enough, this is not any different from ISO/IEC’s directives which read in Part 1:

It is the responsibility of the chairman […] to judge whether there is sufficient support bearing in mind the definition of consensus given in ISO/IEC Guide 2:1996.

“consensus: General agreement, characterized by the absence of sustained opposition to
substantial issues by any important part of the concerned interests and by a process that
involves seeking to take into account the views of all parties concerned and to reconcile
any conflicting arguments.
NOTE Consensus need not imply unanimity.”

The way the “absence of sustained opposition” is typically assessed at W3C is by simply asking whether everybody can live with the proposed decision. As I indicated before asking this question can sometimes lead to a completely different decision. I’ve witnessed it myself. So, this is not just some sort of polite gesture, it is at the very heart of consensus building.

I wish that rule were more broadly adopted. For instance, to my knowledge this was never asked in any way at the BRM for OOXML.

Instead, the BRM proceeded by making its decisions by simple majority vote, leaving no room “to reconcile any conflicting arguments”. In fact, as it’s well known, the vast majority of the issues were “addressed” in bulk via a simple majority vote. So much for ISO/IEC directives. It seems to me that the chair failed to “bear in mind the definition of consensus” or was instructed to forgo that golden rule for the sake of expediency.

As if that was not enough ISO/IEC officials went on to then recommend to its respective boards of directors (SMB & TMB) to dismiss the appeals filed by Brazil, India, South Africa, and Venezuela. Now, I don’t know what an appeal is if it’s not a clear expression of “sustained opposition”.

You would think that given ISO/IEC’s own directives, its officials would have recommended to create a conciliation panel “to take into account the views of all parties concerned and to reconcile any conflicting arguments.” Looks to me like ISO/IEC officials, who have for mission to ensure the directives are followed mind you, have completely lost sight of this fundamental principal. Either that or they should revise their directives to acknowledge that decisions are not actually made by consensus.

As Wikipedia puts it: Consensus decision-making is a group decision making process that not only seeks the agreement of most participants, but also to resolve or mitigate the objections of the minority to achieve the most agreeable decision.”

We’d all be better off if everybody were “to bear that in mind”.

Advertisements

August 7, 2008 - Posted by | standards | , ,

23 Comments »

  1. Hi Arnaud,

    At the BRM “consensus” was dealt with as “absence of sustained opposition” as you note above. However, it was carried out a bit differently than in W3C (the way you describe it). “Consensus” was pragmatically executed as “no more than a couple of countries opposing a resolution”. If my memory serves me correctly, the line between “consensus” and “not consensus” was roughly defined as “maximum of three” countries opposing. I don’t remember any resolution being passed with more than three countries opposing (you can check the meeting report, if you like).

    [quote]Instead, the BRM proceeded by making its decisions by simple majority vote, leaving no room “to reconcile any conflicting arguments”.[/quote]
    That is simply not true for the questions that were dealt with “on the floor”. The very reason we had so little progress the first two days (and the last two days as well) was, that (almost) each original proposal was challenged by other countries and they then went “off-line” to work out their quarrels.

    The very reason a lot of us attending the BRM were so completely trashed on Saturday was that we literally spent every break, lunch-our, evening and sometimes night-hours as well to talk to other NBs about proposals and “reconcile any conflicting arguments”.

    :o)

    Comment by Jesper Lund Stocholm | August 7, 2008 | Reply

  2. According to ISO/IEC the “total of responses addressed at the BRM is 189 – 18.4%”, while the “total of responses decided by group-voting is 838 – 81.6%”.

    Besides, other BRM participants have reported that while they indeed spent a lot of time outside the BRM to come up with proposals they never had a chance to present them.

    Comment by Arnaud Le Hors | August 8, 2008 | Reply

  3. Hi Arnaud,

    “According to ISO/IEC the “total of responses addressed at the BRM is 189 – 18.4%”, while the “total of responses decided by group-voting is 838 – 81.6%”.”

    Yes – that was what the BRM decided to do. As I am sure you remember, not a single country voted against that decision. I am pretty sure some countries thought the vote was a sure key to dismissal of the responses from ECMA, but as it turned out, they were wrong.

    “Besides, other BRM participants have reported that while they indeed spent a lot of time outside the BRM to come up with proposals they never had a chance to present them.”

    Yes, I was amongst these. I worked on some suggestions that was put on the floor and I worked on some that wasn’t. That’s the name of the game.

    Comment by Jesper Lund Stocholm | August 10, 2008 | Reply

  4. That’s the name of ISO’s game, not of consensus building. My point exactly.

    Comment by Arnaud Le Hors | August 10, 2008 | Reply

  5. Hi Arnaud,

    “That’s the name of ISO’s game, not of consensus building. My point exactly.”

    Yes, well truth is indeed in the eyes of the beholder.

    :o)

    Comment by Jesper Lund Stocholm | August 11, 2008 | Reply

  6. What you want to believe as being the truth is certainly up to you but the facts that you actually confirm speak for themselves when it comes to the point at hand.

    For one thing, consensus building takes time, and if there is one thing nobody has had all along in the processing of OOXML is time.

    I suggest you read this other blog entry I found on consensus I found after I posted mine.

    The author gives a striking example of somebody claiming consensus when there is not, as well as some key elements of what truly constitutes consensus building.

    Comment by Arnaud Le Hors | August 11, 2008 | Reply

  7. Hi Arnaud,

    It is true that consensus takes time, but I don’t agree with you that there was not time enough in processing OOXML. Or to put it in another way, if we hadn’t spent so much time talking about stupid things, the time frame would have been highly sufficient. The problem was not the process – it was the people involved.

    The article you found on consensus-building is really good and it applies perfectly to what we are talking about and to what happened at the meeting in Geneva. I don’t remember anyone “strong-arming” everyone into “silence” though, but I do recall some participants that didn’t want to “Move their stake”.

    Oh – and about what I confirm: I haven’t heard anyone claim that the vote on the remaining responses from ECMA (the ones not dealt with on the floor) constituted any kind of consensus – it was just a vote.

    Comment by Jesper Lund Stocholm | August 11, 2008 | Reply

  8. Time? Any NB that felt OOXML had problems that were so severe they couldn’t be dealt with by maintenance could have (and should have) voted reject. However the majority did not.

    It was not MicroSoft steamrollering nor ECMA puppetry nor ISO officials ass-covering nor convenors conniving nor lords a leaping that got IS29500 its accept vote: it was the BRM making enough changes for address enough concerns of enough NBs to change their vote that did it.

    I think the basis of your comments are that consensus cannot operate with a deadline: that when people exhaustively talk through every issue that can result in consensus, but when people prioritize issues and agree to fallback or default positions on issues of lower priority that is not consensus.

    Comment by Rick Jelliffe | August 12, 2008 | Reply

  9. Jesper,
    ISO officials wrote in their documentation sent to TMB and SMB: “The resolutions of the BRM document decisions taken where consensus was reached.”

    As you just said, “it was just a vote” though.

    Rick,
    several countries did vote to reject OOXML but, as you say, the majority did not. That doesn’t make it the consensus though.

    Comment by Arnaud Le Hors | August 12, 2008 | Reply

  10. Rick, if the rules are not applied correctly, then you raise an objection at the time. If your objection does not prevail, then you appeal. To vote against a proposal because of procedural objections, as you suggest, is to confuse the substance of the proposal with the process. And to suggest that approval of the substance of the proposal also denotes approval of the process, that is to say that the ends justify the means, and is particularly ill-advised for any legitimate, rule-based organization.

    The fact is that at the BRM objections were not allowed to be voiced. Resolutions were made, the Convenor asked whether there were any objections, NB cards were raised to make objections, but the Convenor frequently declared the motion to carry, without even allowing the NB’s to state the nature of their objection. For procedural objections, the speakers, including Heads of Delegation, were ruled out of order before they could finish their sentence. This is not consensus as ISO defines it, where consensus requires “seeking to take into account the views of all parties concerned and to reconcile any conflicting arguments.”

    Since almost every Ecma proposal that was discussed at the BRM was found to be defective and require additional changes, it is lamentable that so many other Ecma proposals were approved via a vote, without consensus or opportunity for NB’s to express objections. Remember, the vote to approve the batch of Ecma resolutions was by a very slim margin for many resolutions, e.g. 11-9, and by no stretch of the imagination could constitute “consensus”, especially when the minority view on any resolution was not allowed to speak.

    This was at best a dictated consensus. If no one is allowed to state their objections, then of course there is no sustained objection. This logic might fly in North Korea, but it should not prevail in JTC1.

    Comment by Rob Weir | August 12, 2008 | Reply

  11. @Arnaud

    Tsk. I come back from a nice relaxing holiday to find offensive rubbish like this. You write “It seems to me that the chair failed to ‘bear in mind the definition of consensus’ or was instructed to forgo that golden rule for the sake of expediency.”

    If you look at the correct part of the Directives (relating to Fast Track BRMs) you can read (13.8) “[a]t the ballot resolution group meeting, decisions should be reached preferably by consensus. If a vote is unavoidable the vote of the NBs will be taken […]”.

    To state the obvious, given the time limit for the BRM a vote was “unavoidable”.

    And ultimately of course most big decisions in JTC 1 are taken by voting (letter ballot or in-committee voting). Your suggestion that consensus needs to be the mechanism of all decision-taking is just wrong (even if it is to be encouraged at earlier levels of decision).

    @Rob Weir

    You write “NB cards were raised to make objections, but the Convenor frequently declared the motion to carry, without even allowing the NB’s to state the nature of their objection”.

    “Frequently”? I think you mis-remember.

    On the occasions where there was not unanimous consensus the objector was frequently invited to state their case (when not already aired, or when the topic had not already been discussed at length), often leading to further discussion. The ONE resolution vote where I did not allow discussion was on the editorial batch (resolution 36), where I had stated the day before that there had been enough discussion on this item. This is where a judgement about what constitutes “substantial issues” needed to be taken by the Chair. In my view the limited time could be better spent on bigger topics than this.

    As for not allowing “out of order” discussion — Heads of Delegation had been told what was out of order and that they would be ruled out of order if they tried to go there.

    The view that consensus is not reached if out-of-order topics are not resolved is a recipe for chaos.

    And bringing in North Korea … really!

    – Alex

    Comment by Alex Brown | August 12, 2008 | Reply

  12. Alex,
    you just explained what I’m only trying to say. ISO’s claim that the BRM decisions were consensus based is bogus. Again, that’s what they wrote in the document sent to their board.

    Voting may have indeed be “unavoidable” but then ISO shouldn’t claim otherwise. They can’t have it both ways.

    And, again, recommending to simply discard appeals without giving NBs a chance to express their disagreement is counter to consensus building.

    Comment by Arnaud Le Hors | August 12, 2008 | Reply

  13. Alex, I’m not relying on memory. I have contemporaneous notes, including in particular a transcript of your statements and actions for all days of the BRM.

    I’m not aware of any rules of order, certainly nothing in JTC1 Directives, that allows a Chair to rule all procedural objections out of order.

    A vote was not unavoidable. There was no formal requirement that all comments be processed in one week.

    Agreeing to an unreasonable schedule, and then using that as a defense of ignoring the requirements for consensus is disingenuous. You are using a fictitious restriction that was not required by the Directives (a one week BRM that addresses all comments as an excuse to not do something else that is required by the Directives(allowing objections to be heard, working toward consensus, avoiding mere vote counting). If we follow your logic, we can eliminate discussion and consensus in any JTC1 meeting by mandating an arbitrarily short meeting. If the OOXML BRM can be limited to one week, then let’s limit other BRM’s to 5 minutes and the SC34 Plenary to 15 minutes. What in principle would forbid that as a means to prevent discussion? The answer of course is that if such abuses took place there would be an appeal. Indeed, and here we stand today.

    A more reasonable reading of the phrase “if a vote is unavoidable” would be that if discussion and debate has not lead to a consensus, and it is evident to the Convenor that reasonable progress toward consensus will not be made with further discussion, then a vote may occur. That is fully in accordance with the letter and spirit of the Directives, and allows the Convenor reasonable discretion. But to have a vote without discussion, or indeed without the opportunity for discussion or objection, and to have NB’s indicate that they have an objection, but not to let them state their objection, that, in my opinion, is gross abuse of the role of the Convenor.

    Comment by Rob Weir | August 12, 2008 | Reply

  14. Rob: You write that “A vote was not unavoidable. There was no formal requirement that all comments be processed in one week.”

    But the option not to vote (to have a paper ballot on accepting the editor’s disposition on issues not otherwise dealt with in the BRM) was explicitly given to the BRM for a vote and rejected: indeed, your own blog records this in your excited entry The Art of Being Mugged: it was option 2, “Anything not discussed is not approved.”

    It was not Alex or the ISO staff who decided to have the vote paper vote, it was the vote of the national bodies. The NBs decided they wanted to have a result in the timeframe permitted, even if some would have preferred more time. It was not the prevention of consensus on particular issues, it was an outcome of a consensus! And it was not a block vote, but a vote by each NB on each outstanding issue.

    Perhaps you should check your contemporaneous notes more. It seems the recollections from OOXML-hostile people from some delegations grouped together on the left side of the room (facing the front) bears no resemblance to the experience of us on the right side of the room, to the extent that it seems you participated in a completely different, fabricated meeting.

    I think Arnaud’s blog above is predicated on the wrong object of consensus.–

    The outcome of a BRM is an improved text. So any consensus at the BRM was that some proposed resolution improved the text. The purpose of the BRM was not to produce a perfect standard in which no-one had any outstanding issues, that would be marvelous but patently impossible for a contentious standard. The post-BRM NB votes decide whether the improved text fixed enough to be acceptable to the NBs and that the remaining shortcomings were appropriate for the maintenance process.

    That every issue was voted on, by NBs who were supposed to have arrived with a knowledge of the NB positions (yes, no, abstain, concur, discretion, etc) on each, and that the BRM was given the opportunity to only vote on issues that got discussion time shows there was was indeed consensus sought and arrived for the limited outcome of the BRM.

    Comment by Rick Jelliffe | August 13, 2008 | Reply

  15. […] English terminology, just as they once did with “contradiction”. Now it’s “consensus” that they try to warp. Over the last year I’ve noticed that quite a few people are using the […]

    Pingback by Boycott Novell » Microsoft Insults the Intelligence of Malaysians, Choice of ODF | August 13, 2008 | Reply

  16. Rick,
    what you’re saying is that the BRM chose to vote therefore the result of the vote is the consensus. It’s not.

    Even if we all agreed the sky is green it wouldn’t make it so.

    If the BRM agreed to give up on building consensus and to revert on voting instead then the result is just that: a vote.

    Comment by Arnaud Le Hors | August 13, 2008 | Reply

  17. Consensus to have a vote is not the same as consensus on the matter you are voting on.

    An example that might help crystallize in your mind — there is a consensus in the US that we should have a presidential election in November. But there is not a consensus on who should be elected.

    This should be obvious.

    The consensus requirements of JTC1 Directives apply to the work of the BRM, which is determining the editing instructions that will lead to the Final DIS text. The fact is that there was no consensus on the editing instructions, and that objections related to these editing instructions were passed over by the Convenor at the BRM, and discussion of the vast majority of the proposed dispositions was prevented.

    Comment by Rob Weir | August 13, 2008 | Reply

  18. @Arnaud

    You are selectively quoting the Secretary Generals’ document in a way which leads to a very misleading impression. That document does refer to the resolutions “where consensus was reached” but ALSO to “well-defined voting procedures accepted by the NBs”. Both methods were used at the BRM.

    You entire blog post is predicated on the double straw man that JTC 1 decision making has to be exclusively consensus based (it does not), and that the Secretary Generals claimed the BRM was exclusively consensus based (they have not).

    @Rob

    At the BRM Heads of Delegation were informed they should raise procedural points with the ITTF representatives who made themselves available for this out of session over the course of the meeting, and who were perfectly equipped for that task. There is nothing in the Directives to mandate that a Chair needs to field points of order. In my judgement session time was better spent on technical substance.

    At the BRM there *was* a requirement that business should be concluded in one week, as this time limit had been set by an ITTF ruling and announced to NBs in November 2007. It was the NBs themselves who chose voting as the best tactic given this time limit (though voting took place out-of-session).

    You are quite correct, however, that in general a JTC 1 meeting may not be time limited.

    I agree with what I take to be your general point that a problem with the Fast Track process (and this applies to PAS too) is that a need for consensus is encroached upon by a need for speed. The result of the process is that a limited amount of time results in a certain text – NBs then have to decide whether they’ve been convinced or not.

    As I have written before, my personal opinion is that PAS and Fast Track standardisation processes should be removed from the JTC 1 Directives (or re-thought comprehensively).

    – Alex.

    Comment by Alex Brown | August 13, 2008 | Reply

  19. Alex,

    The quote from the ISO document is actually in response to a complaint that consensus building didn’t take place:

    3. BRM did not take into account NBs’ efforts during the meeting to reach consensus on modifications to the proposed responses from the Project Editor

    3e. Not correct. The resolutions of the BRM document decisions taken where consensus was reached.

    If as you claim this is only meant to say “consensus was reached when consensus was reached” then I claim it’s the above text which is misleading and does not address the point at hand.

    Besides, my post is on the abuse of the term “consensus” by people in general, not just ISO officials. As you can see above, Rick, for one, is still trying to defend that there was consensus…

    Comment by Arnaud Le Hors | August 13, 2008 | Reply

  20. […] Filed under: Uncategorized — ctrambler @ 1:44 pm Arnaud has an interesting discussion about what “Consensus” really mean? I think everyone can in general agree with what “Consensus” really mean. However, we […]

    Pingback by That dirty word “Consensus” « CyberTech Rambler | August 14, 2008 | Reply

  21. The “OOXML Standard” is a purchase made by Microsoft from the ECMA. To look at any deeper is to get into the quagmire of lies, deceits, bribes, etc. spewed out by Microsoft. It is so wrong that it can never be fixed. This whole post is made possible by assuming that Microsoft has some credibility, which it don’t. OOXML is defined as hoax.

    Comment by Dick Kolklayshr | August 14, 2008 | Reply

  22. Alex says:

    “As I have written before, my personal opinion is that PAS and Fast Track standardisation processes should be removed from the JTC 1 Directives (or re-thought comprehensively).”

    I take it he bases this on his recent experience up to, during, and after the BRM. I infer that he should publicly include DIS29500, withdraw his support for it and instead support the appeals against its adoption.

    Comment by Henri Laurie | August 14, 2008 | Reply

  23. I’d just like to say that after reading this page and much of the rest of the OOXML fiasco that many of us out here see the people who were behind pushing this through ISO as people of little to no honesty or honor.

    You can call this a personal attack if you like. I really don’t care. I’m just reporting what many of us judge the behavior to be. We all know what words like “consensus” mean, and they don’t mean things like pushing through a standard of 6500 pages in length in 350 days.

    It’s impossible to fairly cover that amount of material honestly, fairly, and in depth in that kind of time. Other ISO standards processes averaged less than 1/2 page or less of material per day. The OOXML process “covered”, and I use that term very loosely, 18.5 pages per day. The discrepancy in time alone says this process was corrupt to its core, and that those defending such a corrupt process are no better off ethically than the process itself was.

    So, Rick and Alex, just how do you defend covering material 36+ times faster in OOXML standards process than those of SVG, ODF, and others? How can you possibly say this was a fair process or even vaguely in line with how the ISO has handled similar matters in the past? I’m not saying I’ll believe anything you say, I’m just curious to hear how you will twist things.

    Comment by Freddy | August 15, 2008 | Reply


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: