Paul de Laat and Nicholas Munn. Online Encyclopaedias: An Exchange

SERRC —  February 14, 2012 — 3 Comments

Author Information: Paul de Laat, University of Groningen, p.b.de.laat@rug.nl. Nicholas Munn, University of Gothenburg, cma@psy.gu.se.

de Laat, Paul and Nicholas Munn. 2012. “Online Encyclopaedias: An Exchange” Social Epistemology Review and Reply Collective 1 (3): 10-13.

The PDF of the article gives specific page numbers. Shortlink: http://wp.me/p1Bfg0-bO

Please refer to:

Editor’s Note: Conducted over email during January and February 2012, this exchange between Paul de Laat, University of Groningen, and Nicholas Munn, Monash University, addresses issues in, and related to, their articles published in Social Epistemology 26.1 (2012). Both de Laat, in “Open Source Production of Encyclopedias”, and Munn, in “The New Political Blogosphere”, examine questions related to social epistemology and social media. This exchange focuses on online encyclopedias. Please go to “Social Epistemology and Social Media: An Exchange” to read Paul’s questions and observations and Nick’s replies to issues involving social epistemology and social media.

In this exchange, Nicholas Munn (NM) offers questions and observations to which Paul de Laat (PDL) responds.

NM: What is the relative importance of each of the online encyclopaedias examined? Has any one of them clearly been a leader/innovator, such that the others can be seen to adapt to them (i.e. Wikipedia)?

PDL: Wikipedia is the early innovator that adopted the wiki format (allowing everybody unrestricted and immediate access to textual entries). As mentioned in the article, (de Laat, 2012) many others adopted the same format later on. At the same time, they introduced ever more moderation and restrictions on access (in an effort to obtain the guarantees of expertise). Note that Wikipedia, in turn, has been inspired by the open source software movement that also allows access to the source code tree (at least if a contributor has been accepted as “developer”, not just “observer”). Note also, that h2g2 is the odd man out. It also started early on, but in their own ways (without a wiki).

NM: Might be interesting to expand the discussion out to cover other online encyclopaedias that have explicitly rejected the open source model, such as the Stanford Encyclopaedia of Philosophy (SEP). Is there an important difference in the epistemic status of a resource like this? Recalling the now famous argument that Wikipedia is just as good as Britannica (EB), does the same apply here?

PDL: Wikipedia might be of the same quality level as EB; at least if similar (scientific) topics are drawn into the comparison. (Why this might be so is a matter of speculation. I myself am working on the hypothesis that at least some amateurs have an opportunity in Wikipedia to learn and acquire so-called “interactional” expertise.) A whole new discipline is unfolding that compares the quality of outputs from the various encyclopedic models. As for the SEP, Wikipedian articles about philosophy may possibly match their quality. And then again I would speculate that this is so since 1) trained philosophers also provide a helping hand to Wikipedia, and 2) untrained philosophers obtain some training by participating in discussions on talk pages and working in the wiki; so gradually they are also able to contribute quality. But all this is pure speculation.

NM: We see from table 1 (de Laat, 2012, 77) that the other wiki based encyclopaedias all began much later than Wikipedia. To what extent do you believe that the differences in editorial style embraced by these other projects reflect an attempt to avoid ‘mistakes’ made by Wikipedia, or, if not mistakes, at least certain outcomes considered undesirable?

PDL: It is interesting to note that many encyclopedias in my sample distance themselves from Wikipedia explicitly (“we are not Wikipedia”). Citizendium advertises itself as Wikipedia “corrected” on 2 counts: 1) an account with one’s real name is required; and 2) experts are appointed as “topic editors” to oversee wiki developments in the entries that belong to their particular topic (and to prevent escalating “edit wars”). Scholarpedia and the Encyclopedia of Earth also employ a wiki — but in more restricted ways. Often only accepted experts may participate; moreover, participation can be on invitation only. So the whole idea of crowd participation is obviously undermined; the wiki turns into a content management tool.

NM: The wikipedian assumption of trust seems to work for many issues. That is, Wikipedia is often an excellent source for particular, non-controversial details. However, the assumption of trust also seems to breakdown in marginal cases, those where, for whatever reason, sharply opposed positions occur amongst the (potential) contributors. In such circumstances (evolution, abortion, various nationalisms), the wikipedian approach has obvious flaws …

PDL: Which flaws are you referring to? I only know that Wikipedia takes pains with such entries to keep them under control. All kinds of people are continuously watching the entries as they develop; both new edits and new entries are carefully patrolled. Banners often announce that this is a heated subject, that contributors should take care, that vandalist edits are likely to be deleted very quickly, that they risk to be banned, and so on and so forth. As a rule biographies of living persons in particular are closely watched. Take the entry on Obama. It is on ‘article probation’, meaning that disturbances and obnoxious behaviour are taken even more seriously than elsewhere, the standards of behaviour to be adhered to are much stricter than normal. If the lines are crossed just a little bit this will quickly result in banning users for the topic involved for a period of time. Or take cold fusion: it is placed under a regime of ‘disciplinary sanctions’, meaning (in a similar fashion) that the rules of proper Wikipedian behaviour are closely watched and transgressions strictly sanctioned. And take global warming: editing the entry is forbidden until tempers have calmed down (the page is ‘protected’; in this case ‘semi-protection’, meaning that experienced registered users may still edit).

NM: Given both the array of proposals for change in wikipedian practices and the apparent unwillingness of Wikipedia itself to embrace these changes, do you see a place for a new online encyclopaedia to pick up the basic strands of wikipedian ideas, and implement them in a way that avoids the issues Wikipedia has? Citizendium explicitly started with this idea, but seems to represent a rejection of the wikipedian approach rather than a modification thereof.

PDL: Citizendium cannot be interpreted as rejecting a Wikipedian approach: they endorse the wiki format, though with some extras (real names and topic editors). With this slightly modified approach they did not meet with a lot of success: the number of entries (about one percent of the Wikipedia total) and more in particular their growth rate are not spectacular, the enterprise seems to be stalling by now. Obviously, it is not easy to find a place under the sun when Wikipedia is around—an obstacle that any new encyclopedic venture will have to overcome. So in the end, I do not see any viable way of starting a new open-content encyclopedia in general. Wikipedia has simply grown too large. For niche topics some room may exist.

NM: Do reputation scores (de Laat 2012, 88-89) lead to unacceptably high barriers to entry for new participants in encyclopaedia projects? Is a new commentator to be treated initially as neutral, or do they have significant disadvantages to overcome simply in virtue of not having previously participated in the project? A possible solution is connecting edits to some ur-identity for each user, that runs over a range of areas. However, suggestions like this (for example, the increasing use of Facebook profiles for commenting on blog articles) have their own downsides, such as the apparent harms to women participants that result from abuse when the identity of the participant is known.

PDL: This has to be a misunderstanding. If the dynamic reputation model (de Laat 2012, 97-99) is introduced, newcomers by default obtain a low reputation score. This score then grows or diminishes in relation to the longevity of all edits contributed. Those who contribute many edits that stand the test of time see their reputation rise; those who contribute edits that soon perish (being deleted by co-contributors) will see their reputation sink. Nevertheless, there is no barrier to entry: anybody can start without any obstacles. But let us say the right to vote on proposals, to review new edits coming in, and so on, will only accrue to those with a high enough reputation. And vice versa, sinking reputation will imply closer scrutiny of one’s new edits than for average Wikipedians. The solution to seek guarantees in real-life identities (of a kind) has always been rejected by Wikipedia, since they feel that it may deter those contributors that prefer to remain anonymous but have much to contribute in spite of this. Not all ‘anons’ are vandals (an estimate is: only 20% of them)!

NM: Wikipedia suffers significantly from a lack of assurance, as set out on pages 92-94 (de Laat 2012). However, it seems as though a combination of ubiquity/size and multiple redundancy (in that for any topic included in the wiki, there are multiple willing contributors, who tend in time to balance their opinions/perspectives out), this flaw is in part addressed. As such, could one model the importance of the assurance standard as a function of the size and/or number of contributors to the wiki itself?

PDL: Possibly a model runs like this. Lack of assurances as associated with anons is mainly a problem in the short run: malevolent anons tend to enter the scene and leave. So whenever an entry is populated with at least some veterans that guard the gates, the vandal problem can be held in check. And that is pattern one sees often: a core of a handful of veteran Wikipedians (some 10% of all) are involved in an entry over a longer time and take the lead as to where the entry is heading. All along, they police unconstructive edits.

NM: The expertise condition seems vulnerable to the flaws mentioned regarding reputation scores above. Especially in light of the notoriously hostile environment of the internet to minority/disadvantaged contributors, the exclusion of legitimate sources of information through such entry requirements seems a real concern. If, as you suspect, an expertise-based approach is beginning to take hold in Wikipedia, I would be concerned that there will be a real loss of perspective entailed by this change.

PDL: Wikipedia is taking pains to combine the 2 perspectives. On the one hand, expertise (edit longevity, de Laat 2012, 96) may become important to calibrate the quality of entries (and stimulate further development). On the other hand, this is not to mean that anyone is to be excluded as source of information. Barriers to entry are and probably will remain very low (provided you have an Internet connection and speak an appropriate language). So what is the loss of perspective you are afraid of?

NM: The following passage “Finally, it has to be observed that in case the real contributory experts stay away from some Wikipedian areas, contributors with high edit longevity may only seemingly qualify as interactional experts in them.” (de Laat 2012, 97) brings to mind a comment I believe (but am not sure) David Chalmers made regarding Wikipedia, namely that he had given up on attempting to address certain flaws in its coverage of his areas of philosophical interest, because his alterations were being reversed on the grounds that Wikipedia did not consider him an authoritative source. (I could send him an email and check if I am recalling the particulars of this story correctly, if you would like?)

PDL: Authority obviously is a big problem in Wikipedia. Sanger (who founded Citizendium) is fond of saying that in the worst case, edit wars erupt causing entries to fluctuate around an equilibrium – but never reaching it entirely. I have no real answer to that — except for arguing that also in the sciences, editing conflicts often have no natural resolution and linger on. About climate change we still have wars going on, in spite of all the ‘real’ experts involved. So maybe Wikipedia is just reproducing a more general phenomenon.

Contact details:
p.b.de.laat@rug.nl
nicholas.munn@monash.edu

References
De Laat, P. 2012. Open Source Production of Encyclopedias: Editorial Policies at the Intersection of Organizational and Epistemological Trust. Social Epistemology 26: 71-103.

3 responses to Paul de Laat and Nicholas Munn. Online Encyclopaedias: An Exchange

  1. 

    Hi Paul,

    A follow-up to the discussion about reputation scores as a potential means of avoiding some of the problems associated with open source editing of this nature: I asked “Do reputation scores (de Laat 2012, 88-89) lead to unacceptably high barriers to entry for new participants in encyclopaedia projects?”, I do not think that in asking this I misunderstand your position. However, my question could have been explained more clearly. In response, you say “Nevertheless, there is no barrier to entry: anybody can start without any obstacles.”

    Now, what I was trying to get at was the idea that the very existence of a reputation score may over time create barriers to entry for new participants in an encyclopaedia project. If the score is not zero-sum, then over time the score of long term (positive) contributors will steadily rise, and the expectations regarding a ‘good’ reputation score will similarly rise. This could lead to a situation where a ‘good’ reputation score is 100, while a new contributor starts at 0, and must amass a significant body of edits to reach 100, and to be considered a valuable participant. In such a case, being late to join the project seems to itself be a disadvantage, or barrier to acceptance as a full member. You are disadvantaged particularly because you have yet to do enough, rather than because of the content of your actions.

    A parallel could be found in the difficulty new members have of integrating into existing communities of any kind. (Whether this is in immigration to a new country, or joining an on-line community or game). If there is a measure by which the established and new participants can be differentiated (language proficiency, number of posts, player level in an on-line game), the new participants are disadvantaged. Dynamic reputation scores seem to provide such a measure.

    • 

      Hello Nicholas,

      I understand your concerns now. So very good that you have clarified your position! But the ”reputation model” is not so overwhelming and oppressing as you depict it. For one thing, having a low score (as a newcomer) does not hinder one to contribute; only privileges do not accrue. For another, the model does not use a steadily rising score; on purpose, there is an upper bound, a ceiling. Having reached the ceiling, any more lasting edits do no longer contribute; but any of one’s edits being deleted do, by diminishing one’s reputation.

      In spite of all this, you have a point: becoming an accepted and reputable ”expert” seems a huge task to which one would need to devote much time and energy. Over time, that bar seems to have been raised higher and higher. When Wikipedia started, newcomers were among equals only. By now, there is also a class far above them that scrutinizes every new edit and every new article. These patrollers may become quite strict, even nasty. That such an extra layer creates resentment emerges clearly from the enduring discussions about introducing a system of reviewing new edits (for vandalism) BEFORE being accepted in the public version of an entry (system of review, flagged revisions). Rights of review then are granted only to more experienced Wikipedians, creating a new class of its own. At times these discussions have been very heated, and full of anger springing from the perception that another step towards bureaucracy would be taken, creating a society of 2 classes, the experienced Wikipedians vs. the newcomers. [Of course all this is shameless self-promotion, as I just published an article about this topic in Ethics and Information Technology, 2012, online first..]

      So I acknowledge your fears.

Trackbacks and Pingbacks:

  1. Nicholas Munn and Paul de Laat. Social Epistemology and Social Media: An Exchange « Social Epistemology Review and Reply Collective - February 14, 2012

    […] to social epistemology and social media. This exchange focuses on political blogs. Please go to “Online Encyclopaedias: An Exchange” to read Nick’s questions and observations and Paul’s replies to issues involving online […]

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