[nSLUG] Complete control is better? (was: Wikipedia)

Daniel Morrison draker at gmail.com
Wed Nov 4 12:05:45 AST 2009


Sorry, Richard, but I disagree with you on pretty much all counts. In
particular, your assertion that exploitable flaws in security software
are "immediately evident" is entirely wrong.

> Who is actually checking that level of expertise [at Wikipedia] versus who is checking it at Britannica?

As Ben pointed out, it's everybody/anybody, by using critical
thinking. A diploma does not guarantee correctness. Peer review does a
much better job.

You worry that too many people in the world will take Wikipedia (or
Britannica, or any other resource) at face value without engaging in
critical thinking. But this is not a flaw in Wikipedia. Our failure to
teach people to question authority will not be solved by installing
additional authorities.

Also:

> ***   True, David, and unfortunately it's becoming more true as
> literacy and journalism standards continue to fall. That aside,
> though, until the standards lowered, I trusted printed media well
> over on-line media that is publically editable.

I don't believe that literacy and journalism standards are "continuing
to fall". Literacy, in particular in the long term, is up. As for
journalism -- sensationalist crime and gore on inky broadsheets were
the norm in Victorian England. I don't think the 21st century has a
particular claim on a lack of journalistic integrity. While the rise
of individual publication on the Internet has certainly had an effect,
I think there are merits to the argument that it has permitted more
"calling to account" of questionable journalism than not.

I don't trust any media due to its format (printed vs. online). I base
my trust on media by the following criteria:
- how well they support their positions using logical reasoning
- how reputable they are considered by their peers (and mine)
- how relevant and reputable are considered the sources they cite
- how openly they admit to their own bias

Now, Wikipedia is a special case, as every page may have many authors.
Thus it is more difficult to apply item #2. However the other items
still apply.

> Until corrections happen quickly and accurately,

Have you actually got any statistics on how quickly and accurately
problems with Wikipedia are corrected? I think you might be surprised.

Bear in mind that there are two types of problems:
- widely held but incorrect assertions of fact. These are called
'mistakes', and they affect Britannica also. Wikipedia has been shown
to be on par with (or at least not outrageously worse than) Britannica
in this respect, but "new editions" of Wikipedia can be published
instantly. Not so with Britannica.
- vandalism. Vandalism to Wikipedia pages is typically dealt with very
promptly. I'm talking about in the order of minutes, maybe hours at
most for science content. Biographies of celebrities, I'll admit, are
probably a lot murkier.

One last note: you keep talking about having to trust "the general
public" with regards to information on Wikipedia. This is a
generalization which is incorrect. For any topic on Wikipedia, it is
not "the general public" who edit, making up whatever they feel sounds
good. People who have an interest in a topic are those who make the
edits. It's a self-selecting system, and the "peer review" is made by
other people who have similar interests.

I guarantee, if you select a topic you know nothing about, and start
to make changes in an indiscriminate manner, you will find plenty of
irate wikipedians reverting your changes in short order.

-D.

2009/11/4 Richard Bonner <ak621 at chebucto.ns.ca>:
>
> On Tue, 3 Nov 2009, Daniel Morrison wrote:
>
>> 2009/11/3 Richard Bonner <ak621 at chebucto.ns.ca>:
>>>
>>> On Mon, 2 Nov 2009, Ben Armstrong wrote:
>>>>
>>>> On Mon, 2 Nov 2009 12:23:03 -0400 (AST)
>>>> Richard Bonner <ak621 at chebucto.ns.ca> wrote:
>>>>>
>>>>>    Until they get complete control over it, there's a good reason I
>>>>> call it "Why? Ki-pedia".
>>>>
>>>> So when large, monopolistic software companies exercise "complete
>>>> control" they're evil, but when encyclopedists do, it's good?
>>>
>>> ***   Hmm, I should clarify...  (-:  What I meant by "control" was
>>> that editors would need to verify information as they would with a
>>> printed encyclopedia.
>>
>> It sounds the same to me...
>
> ***   Similar, but not the same in that the editors should be some calibre
> above the general public. I don't think Britannica hires persons off the
> street, so to speak.
>
>
>>>> Why do you hate freedom? :)
>>>
>>> ***   No hate for freedom, Ben;
>>
>> I'm quite certain this question was asked tongue-in-cheek...
>
> ***   Oh, sorry - I guess that went over my head.  (-:
>
>
>>> I am just against the abuses and
>>> inaccuracies that permeate public resources when there are no
>>> overseers.
>>
>> I am against the abuses and inaccuracies that permeate
>> privately-controlled resources when there are overseers who exercise
>> "complete control".
>
> ***   I would still trust them over the general public. There is some
> minimum level of education required for that job at Britannica.
>
>
>> You see, there will be abuses and inaccuracies either way. There is no
>> free lunch, the world is not safe, we cannot put 100% trust in
>> anything, etc.
>
> ***   Oh, absolutely. However, there should be some minimum beyond what
> persons say their credentials are when submitting to an on-line resource.
> Who is actually checking that level of expertise versus who is checking it
> at Britannica?
>
>
>> The question to ask is: having understood that there WILL be abuses and
>> inaccuracies, would we rather allow those abuses to be served to us in a
>> top-down way by self-proclaimed authorities, with no recourse to correct
>> those abuses when they are noticed by that minority who can correct them? Or
>> would we prefer an open, bottom-up system in which individuals can exercise
>> the power to correct flaws in a collaborative way?
>
> ***   I'd want some of each.
>
>
>> The argument has been made and is done with for security concerns -- the
>> experts consensus is that the open method is better. Security through
>> obscurity does not work. Security through publishing your algorithm,
>> essentially challenging the world to break it, does work. This is why ssh is
>> secure, and Linux is secure, and DVD encryption and MS Windows is not.
>
> ***   I don't think that is a perfect analogy because when software doesn't
> work it is immediately evident; it gets corrected or people move on. With
> information accuracy, it is not evident except to those that know better.
> Too many of those that do not, will use it as though it is, and then
> perpetuate it. This has been shown to be true through the popular-media
> examples I gave.
>
>
>> We might still spend a lot of time debating this, but to my mind, there is
>> a quick way to short-circuit the problem to an answer. There is another
>> factor: that such large and comprehensive systems, as they grow, become TOO
>> large and TOO difficult for a single entity to effectively control. A small
>> number of coders CANNOT verify every line of code in a large closed source
>> software, and we have WIndows. But a large number of coders can do such a
>> thing in a large open source software, and we have Linux.
>
> ***   Yes, but that is assuming they know there is a problem and that
> someone actually spends the time to correct it. Until corrections happen
> quickly and accurately, Why?-Ki cannot reach its full potential.
>
>   Now of course, it's in its growing stage. It will be interesting to see it
> in five years. In the meantime, I have seen to many errors, biases and
> misconceptions for my liking.
>
>
>> A small number of overseers at Britannica CANNOT effectively oversee and
>> keep up to date a large body of work such as an encyclopedia.
>
> ***   That is why it takes so long between editions. The Oxford Dictionary
> just went through exactly that in recent years.
>
>
>> A large number of overseers CAN do so at Wikipedia.
>>
>> -D.
>
> ***   Provided they are accurate and actually spend the time to do so, I
> agree.
>
> Richard
> _______________________________________________
> nSLUG mailing list
> nSLUG at nslug.ns.ca
> http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug
>
>



More information about the nSLUG mailing list