The Wikipedia: Knowledge as social, fallible, externalist and holistic
The Wikipedia: Knowledge as social, fallible, externalist and holistic


Knowledge has often been conceived in an individualistic, internalist and intellectualist manner. In this paper we use the example of the Wikipedia and other Internet resources to argue for an alternative, social, fallibilist, externalist and holistic conception. We analyze issues concerning the reliability of publicly constructed bodies of information and extract conclusions for some central debates in epistemology, such as the sceptical challenge, the sources of normative evaluation, the role played by the community and the world in the constitution of knowledge or the intelligibility of a radical mistaken thinker. The paper finishes by offering some connections between these epistemological consequences and ideas found in the philosophy of Socrates, Davidson and Wittgenstein.

Table of contents

    Most traditional accounts of knowledge are individualistic, internalist and intellectualist. Knowledge is attributed to an individual human being (and, perhaps, also to animals or to complex mechanisms, but we will leave this question behind), who knows what she knows and what she ignores, where knowledge is taken to be basically propositional in character. The emphasis on an individual who, transparently, knows facts makes a mystery out of many ordinary uses of the verb “to know”. Let’s think of the following situation: a group of friends are travelling from a city to another and they must get to a certain concert hall. One of them can find her way around the city quite well and has a clear idea as to the location, but has never seen the hall; another one has been driven to the hall in previous occasions but didn’t pay much attention to the streets that lead to it. None of these two can drive. The third one is unfamiliar with the city but has driven to it in the past. They know how to get there, but their knowledge is far from propositional and none of them has it (and, if they don’t trust each other’s sense of directions very much, none of them could be said to know that they know). This phenomenon replicates no matter where we look. Human beings have managed to build airplanes or television sets, but it is doubtful whether any single person would know how to make one from scratch. No single human being can be praised for discovering the right time of year to harvest, to prune trees, or to choose the right seeds to plant. Even in cases that seem paradigmatic of individual abilities—say, write songs—, we find important examples where those abilities are spread across individuals—think of the standard division of labour between lyricist and composer in opera or in the American Songbook, but also on the abyss between Lennon’s and McCartney’s joint compositions for the Beatles versus their solo careers. Sometimes it takes two to know, sometimes it takes the whole village.

    Even though we find the internalist and intellectualist commitments of traditional epistemology seriously questionable, in this paper we won’t dwell on them and instead will concentrate on the shortcomings of individualism.1 To do this we will examine the epistemological status of encyclopaedias and dictionaries and argue that the development of the Internet and the intrinsically collective nature of its use makes it even clearer that individualistic conceptions of knowledge are mistaken. We will focus on the stunning example of the Wikipedia, but will say something about dictionaries first.

    In order to introduce the issues concerning the reliability, accuracy and breath of scope of collectively built bodies of knowledge, we’d like to illustrate the intuition behind our reasoning by reference to judgements about the correctness in spelling and grammar. Languages that bear the dubious blessing of a Royal Academy (as Spanish or French) have the advantage of counting with a committee of experts, often composed of some of the most prestigious (generally male) representatives of the language (writers, linguists, journalists...) dedicated to their study and preservation, but at the same time these academies are given enormous normative powers which tend to be used in a rather conservative manner. Speakers have no saying on the election of the members, which gives rise to elitism, a variety of the intellectualism we mentioned at the beginning of the paper. In contrast, other linguistic traditions—such as the one represented by the English language—lack an ultimate authority on correctness, and this situation forces its lexicographers and grammarians to be much more up to date with the evolution of their language (an amusing example of this is the 19th Century grammatical rule “Never use a preposition to finish a sentence with”). Perhaps the most unfortunate consequence of this academicism is that a great number of speakers, generally those from more humble backgrounds, are dispossessed even of the right of being authorities in their own language. In opposition to this we feel compelled to side with Davidson in the claim that, in an important sense, a native cannot make mistakes in her own language—the idea being that a language should be spoken the way natives do (Davidson 1986)—, or with McDowell’s reading of Wittgenstein’s rule-following considerations (McDowell 1984)—there cannot be a general interpretation or a ready made universal rule with higher normative status than the practices under evaluation themselves. Dictionaries can, at their best, tell us what is generally done with the words of a language. Before the Internet there was no easy way to count uses; now any search engine would give you a very good idea of how people generally spell words (for instance, our text editor spelling tool does not approve of “encyclopedia”, but 21 million entries found by Google versus a little more than 1 million for “encyclopaedia” should suffice to also accept the former spelling). We feel that this casts doubts on the very idea of a principled divide between the description of practices and the institution of norms and will briefly come back to the issue with regard to the Wikipedia.

    A few months ago we read an article on a webpage that included the following caption: “This article is taken from the Wikipedia. The Wikipedia is an encyclopaedia freely written and revised by the users and any entry may include inaccuracies or factual errors. However, independent studies show that it is a highly reliable source of information.” One of these studies was commissioned and published by the prestigious scientific journal Nature (15-12-2005), where 50 articles on scientific subjects from the Wikipedia and the Encyclopaedia Britannica were peer reviewed and a very similar number of serious and of minor errors were found in both publications.2 Everyone is allowed to edit, a policy that follows from the explicit assumption of good faith (but good faith needs not be assumed in the presence of evidence to the contrary). Someone unfamiliar with the idea would think that the vandals would dominate. There are important correcting mechanisms: previous versions of the articles can be consulted, what allows readers and editors to recover good material that has been replaced by less accurate, or more biased texts; the Wikipedia is not a democracy and it distinguishes up to five different levels of privilege to users; users may be temporarily or permanently banned; some especially polemic entries are protected or semi-protected, etc. However, these mechanisms do not preclude that a user that registered as little as five days ago may change the title of page by moving it or edit a semi-protected page. These regulations and hierarchies are insufficient to grant an individualistic and elitist conception of knowledge.

    Furthermore, besides the open question regarding the comparative frequency of errors or deliberate boycott, there are numerous areas where the Wikipedia fares better than most traditional encyclopaedias. For instance, it is constantly updated and the range of topics is much broader. More importantly, the decision regarding what portions of knowledge are relevant to deserve an entry or to appear within entries does not depend solely on the imagination and judgement of a person or small group of persons but it is open to a large group of collaborators, within the limits imposed by the control mechanisms discussed in the previous paragraph and by some hierarchical restrictions regarding what deserves or not an entry. This feature does justice not just to the collective elements of knowledge, but also to its essentially externalist nature: the massive number of contributors, editors and administrators guarantees that many more aspects of reality are embraced by the encyclopaedia.

    The externalist consequences are important, but we also want to draw some conclusions regarding fallibilism, the ineliminable social aspects of knowledge and holism. The idea that any given entry in an encyclopaedia (written by experts or just by anyone willing to do so) may be wrong, as suggested by the quotation in the previous paragraph, can be taken to be food for sceptical worries. The argument is familiar from all forms of scepticism: if any p amongst my beliefs can be wrong, they all can. It can been argued, following Davidson, that the intelligibility of a body of beliefs demands that many of them are correct and shared by the interpreter (in fact, that many of them constitute knowledge). However, the idea that some of our beliefs should be certain and beyond any possibility of revision opens the door for a radically misplaced thinker: if it is conceivable that someone may know everything there is to be known, someone may be absolutely wrong and still count as a thinker (i.e., the world could play absolutely no role on the conditions of possibility of thought). In order to avoid the demand for incorrigibility—and the dogmatism that it invites by making room for an omniscient knower or intuitive intellect—we think that a certain level of caution is necessary regarding anything we think, hear or even see. This should be the lesson to learn from the sceptic. In order to have knowledge we do not need unshakable certainty. Rather the opposite: we can, and must, admit that any piece of information or belief may be wrong in order to make sense of the possibility of knowledge. The sceptical move from “there are mistakes” to “the source is unreliable” would affect not just the status of the Wikipedia as a valuable source of information, but the status of any source or person whatsoever. The Wikipedia policy guidelines insist that the information included should be reliable and verifiable; these are traditional epistemological values but, once again, they both point in the direction of a social understanding of knowledge.

    The second consideration we would like to rehearse concerns the role that the community occupies not only as attributor of knowledge but as knowing subject. Encyclopaedias in general, as collectively built bodies of knowledge, show that the linguistic community plays a role sanctioning what counts as knowledge but also, and more fundamentally, that it is a genuine depository of culturally and historically accumulated knowledge. The Wikipedia makes the case even stronger, given its huge base of contributors. In connection with the fallibilist and externalism defended above (i.e., with the issue of reliability despite the possibility of error), it is important to notice that we are not siding with any strongly naturalistic conception of reliability: the Wikipedia can be seen as a reliable system to generate beliefs and communicate knowledge precisely because of its externalist and social character. However, the analogy with the naturalistic version of reliabilism is useful: our visual system is generally reliable despite the existence of visual illusions.

    We believe that seeing knowledge as social, externalistic and fallible leads to a broadly holistic view. A technical feature of the Wikipedia, shared to a much larger extent with other Internet resources than with traditional encyclopaedias, also points in the direction of such holism. Besides offering external references (often by giving links to sites where texts, music or reproductions of works of art can be found), the internal cross-referencing mechanisms highlight the intimate interdependence among pieces of information.

    Of course, the view of knowledge that we are propounding by means of the example of the Wikipedia and other Internet resources is not new at all and it can be defended independently of any reference to them. We would like to finish this paper by pointing out three notable examples of such a view. The first constitutes a founding insight for Western philosophy, even though it is often ignored: Socrates’s insistence on the essentially dialogical nature of knowledge. A very clear reminder of the centrality of communication for knowledge can be found in Sloterdijk 1988 (see especially chapter 3): he argues that what truly distinguishes Socrates from all other philosophers is his idea that questioning and dialogue should not be seen as a path to constructing philosophical theories, but rather as an acceptance of the philosopher’s (the individual’s) ignorance. As soon as a positive answer is attempted, we move away from the Socratic method into the realm of Platonism.

    The importance of dialogue has been strongly defended by many authors within the hermeneutical tradition, most notably by Gadamer. However, we find Davidson to be its most inspiring champion. His ideas about radical interpretation, where the attribution of meaning must be done on the assumption that speakers and knowers have an ineliminably normative character and that interpretation is done in terms of the specific situation of interpretee and interpreter in their environment (Davidson 1973), his view that linguistic knowledge cannot be separated from general knowledge of the world (Davidson 1986) and his arguments for the idea that triangulation of self-knowledge, knowledge of the world and knowledge of other minds is a precondition for thought (Davidson 1991) say much in favour of the view we are putting forward.

    Even though we lack the space to properly discuss his approach, we cannot ignore the contribution of Wittgenstein to the debate. We have already mentioned that his discussion of rule-following clearly gestures towards a social and externalist conception of thought and language. This discussion sits very well with one the “five pillars that define Wikipedia’s character”: Wikipedia does not have firm rules besides the five pillars (it is an encyclopaedia, has a neutral point of view, is free content that anyone may edit, has a code of conduct, and promotes bold editing, moving and modifying articles). No less important is the idea that the normative aspects of language, knowledge and thought cannot be reduced to something outside the realm of normativity, and must be situated within the realm of forms of life, of socially constituted practices.


    1. Britannica 2006:
    2. Davidson, Donald 1973 “Radical Interpretation”, in: Inquires into Truth and Interpretation, Oxford: Clarendon Press, 1984, 125-139.
    3. Davidson, Donald 1986 “A nice derangement of epitaphs”, in: Truth, Language, and History, Oxford: Clarendon Press, 2005, 89-107.
    4. Davidson, Donald 1991 “Three varieties of knowledge”, in: Subjective, Intersubjective, Objective, Oxford: Oxford University Press, 2001, 205-220.
    5. McDowell, John 1984 “Wittgenstein on following a rule”, in: Mind, Value and Reality, Cambridge, Mass.: Harvard University Press, 1988, 221-62.
    6. Sloterdijk, Peter 1988Zur Welt kommen – Zur Sprachen kommen, Frankfurt am Main: Suhrkamp Verlag.
    “Individualism” is sometimes used as synonymous with “internalism” (Tyler Burge would be a clear case). However, our use of the term is meant to establish a contrast between the individual and the community, not between what is internal and what is external to the individual.
    In March 2006 the editors of the Encyclopaedia Britannica published advertisements in UK and US criticizing the survey and Nature responded, convincingly to our minds. The texts can be found in Britannica 2006 and Nature 2006.
    Manuel Pinedo-García and Cristina Borgoni-Gonçalves. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.