Tyler Burge: Intellectual Norms and Foundations of Mind. In: The Journal of Philosophy 83 (1986) 12, 697-720, hier 707-708.
»We begin by imagining a person A in our community who has a normal mastery of English. A's early instruction in the use of 'sofa' is mostly ostensive, though he picks up the normal truisms. A can use the term reliably. At some point, however, A doubts the truisms and hypothesizes that sofas function not as furnishings to be sat on, but as works of art or religious artifacts. He believes that the usual remarks about the function of sofas conceal, or represent a delusion about, an entirely different practice. A admits that some sofas have been sat upon, but thinks that most sofas would collapse under any considerable weight and denies that sitting is what sofas are preeminently for. A may attack the veridicality of many of our memories of sofas being sat upon, on the grounds that the memories are products of the delusion.
A is willing to test his hypothesis empirically, and the sociological tests he proposes are reasonable. A also offers to demonstrate by experiment how the delusive memories are produced. He is sophisticated, and the tests would require elaborate controls. We can even imagine that the theory is developed so as to be compatible with all past experience that might be thought to have falsified his theory. Thus a normal but sophisticated conception of confirmation accompanies A's unusual theory. We may imagine that if we were to carry out his proposed experiments, A would come to admit that his theory is mistaken.
As a second step, imagine a person B (or A in nonactual circumstances) who is, for all intents and purposes, physically identical to A. He has the same physical dispositions, receives substantially the same physical stimulations, produces the same motions, utters the same sounds. Like A, B hears, though seldom, word forms that are counterparts to the truisms that A hears. But in B's situation, these word forms are not taken as truisms; they are contextually appropriate remarks that do not purport to convey a general meaning. (They could be lies or jokes, but it is more natural to take them as not-completely-general contingent truths.) The objects that B is confronted with are objects that look like sofas, but are, and are widely known to be, works of art or religious artifacts sold in showrooms and displayed in people's houses. Many of these objects would collapse under a person's weight. There are no sofas in B's situation, and the word form 'sofa' does not mean sofa. Call the relevant objects „safos.“ B assumes that most people would take these objects to function primarily as seats and that the remarks he hears are communally accepted truisms. But, like A, B develops doubts. At least by the time B expresses his skepticism and his theory, he is correctly doubting that safos function as furniture to be sat upon. Thus B's thoughts differ from A's. […]
The conclusion is that A and B are physically identical until the time when they express their views. But they have different mental states and events. A has numerous mental events involving the notion of sofa. B's skepticism does not involve thinking of anything as a sofa. I do not assume that what I have said about the case in non-propositional-attitude terms entails that A and B have different thoughts. I think, however, that it is overwhelmingly plausible that they do and, more importantly, that there is no general objection to the natural view that they do.«
Person A lebt in unserer Welt, denkt über Sofas nach, und gewinnt mit der Zeit, auch durch Äußerungen anderer über Sofas, die ihm gegenüber scherzend gemacht werden, die (falsche) Überzeugung, dass Sofas eigentlich Objekte religiöser Verehrung sind, die gar nicht zum Sitzen gemacht sind und sich auch dafür nicht eignen, z.B. würden sie zusammenbrechen, wenn eine erwachsene Person sich draufsetzt.
Person B lebt in einer Welt, die keine Sofas hat, sondern Safos: Objekte, die aussehen wie Sofas, aber tatsächlich religiös verehrt sind und auch nicht zum Sitzen gedacht, sondern z.B. zusammenbrechen, wenn man sich draufsetzt. B hat angenommen, auch durch Äußerungen anderer, die (scherzend) ihm gegenüber gemacht werden, dass Safos nicht verehrt werden, sondern Sitzmöbel sind, korrigiert aber nach und nach seine Überzeugung, dahingehend, dass es sich um Objekte religiöser Verehrung handle, auf denen man nicht sitzen könne.
Person A und B sich physisch gleich, aber sie haben unterschiedliche mentale Zustände, obwohl sie ähnliches zu denken scheinen, nämlich dass man auf den Objekten nicht sitzen kann. B irrt sich zudem über den Sprachgebrauch seiner Sprachgemeinschaft, was Safos angeht, während A darin korrekt ist, aber sich über die Rolle von Sofas in der Welt irrt.
Das Szenario ist den anderen Externalismus-Gedankenexperimenten ähnlich. Wie bei den andern habe ich Mühe, Burges Punkt zu verstehen, da mir die Position, die sie widerlegen soll, nicht klar ist.