ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ר ר ר File: 06-2-94.TXT - 36 KB ר ר ר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ר ר ר Tadeusz Czarnecki, Krakau - Poland ר ר ר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ר ר ר *The other side of understanding* ר ר ר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ר ר ר Abstract: ר ר ר ר THE OTHER SIDE OF UNDERSTANDING ר ר * * * ר ר Epistemic criteria of X necessary truths including a word ר ר "X" are media of understanding of the word "X". A person ר ר who is not acquainted with criteria of X does not ר ר understand the word "X". Nevertheless, when they are ר ר misused, criteria are also responsible for loss of ר ר understanding.In order to reveal the other side of ר ר understanding one should modify the category of ר ר "language game" by /1/ distinguishing between teaching ר ר and playing epistemic language games and /2/ introducing ר ר the additional category of "language as a whole". These ר ר steps allow to interpret phenomena appearing at the ר ר boundaries of language. Some declarative sentences, when ר ר they are viewed from the perspective of teaching, may be ר ר classified as both synthetic a priori and analytic whereas ר ר the same sentences, when they are viewed from the ר ר perspective of playing, are classified as empirical. ר ר Other sentences, when they are viewed from the perspective ר ר of language as a whole, may be classified as these which ר ר lost their empirical character. An analysis of these ר ר phenomena leads to the conclusion that criteria are misused ר ר when one treats them as genuine moves while playing ר ר epistemic language games. A person who states criteria of ר ר X while playing epistemic games with a word "X" proves that ר ר he does not understand the word "X" because the stating of ר ר criteria is a manifestation of the nonsense of playing. ר ר ר ר * * * ר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר ר ר ר Czarnecki, Tadeusz (1994) The other side of understanding; ר ר in: Wittgenstein Studies 2/94, File: 06-2-94; hrsg. von ר ר K.-O. Apel, F. Bצrncke, N. Garver, P. Hacker, R. Haller, ר ר G. Meggle, K. Puhl, Th. Rentsch, A. Roser, ר ר J.G.F. Rothhaupt, J. Schulte, U. Steinvorth, P. Stekeler- ר ר Weithofer, W. Vossenkuhl (3 1/2'' Diskette), ר ר ISBN 3-211-82655-6, ISSN 0943-5727. ר ר ר ררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררררר Despite all the traps that Wittgenstein's philosophy sets on interpreters some rules of interpreting it have been formulated and are recognized as sound. As to general features of the philosophy of understanding interpretations pretending to orthodoxy hold that Wittgenstein's views, in their critical part, are anti-essentialistic and anti-psychological whereas in their constructive part reveal the importance of explanation. The problem to be solved, according to these interpretations, is put by Wittgenstein like this: How can one understand what one still does not understand? Or, to express it in Kantian style: How is understanding possible? In my paper I do not intend to challenge the correctness of this general approach but to show that it is not the only stimulus for Wittgenstein's reasoning. The other, no less natural, is determined by the opposite questions: How can one cease to understand what one has already understood? How is loss of understanding possible? To gain my objective directly I start where some interpreters end*1*. This means I propose no new analysis of, for example, paragraphs 143 - 184 of PHILOSOPHICAL INVESTIGATIONS but concentrate on morals from Wittgenstein's criticism of essentialism. Essentialism contrasts manifestations of understanding with understanding itself and maintains, for example, that someone's actual development of a bit of an algebraic series is only a manifestation of understanding of the series. The essentialist identifies understanding with nothing else but insight into the essence of the series and the essence, respectively, with the algebraic formula of the series. - The infinite development of the series seems for him to be squeezed into the formula. From the moment someone manages to discover the formula - the picture of the formula flashes through his mind - all possible difficulties are overcome and his instant understanding of the series results in the capacity for its infinite development. Wittgenstein is suspicious about this vision and questions he asks are more or less like these: when we attach such importance to the discovery of the formula do we claim that everyone whom we save from the hardship of discovering the formula simply by showing it directly to him will always be able to develop the series? As the answer is in the negative and the direct acquaintance with the formula may be sometimes so obviously unimportant as to cause no expected manifestations at all will we still suggest that the discovery of the formula - the insight into the essence - is the necessary and sufficient condition of understanding? Wittgenstein is aware of the fact that his arguments may seem not convincing enough. Even if the essentialist acknowledges that an isolated fact, the acquaintance with the formula, does not come up to his expectations he may defend his position saying that the only conclusion he is obliged to draw is that the thing he was looking for - the formula - was inappropriately chosen. The failure he faces does not mean he should stop looking for another thing that would be, this time, entirely satisfactory, i. e., would be the real embodiment of the essence of the series. Does, therefore, the essentialist intend to claim - this is the next of Wittgenstein's questions - that someone's acquaintance with the formula is never the factor that really decides about understanding of the series? Is it the case that no one ever develops the series just because he has become aware of the formula? Now, the essentialist may acknowledge that he was a little inconsiderate and may be prone to modify his view as to the importance of the formula. But he does not have to give up his philosophical position even if it seems to be still weaker than before. If on one occasion a fact does not contribute and next time the same fact contributes significantly to understanding of the series this means - he may say - that the fact is essential in certain circumstances, i. e., in as much as it is the decisive component of a configuration of facts. - The awareness of the formula is essential if it is the last link in a chain of previously fulfilled necessary conditions. Reacting to this last response, Wittgenstein returns to a modified version of his former questions: Does the essentialist claim that the moment a set of conditions - a limited configuration of facts, a compound fact - is fulfilled - discovered by someone or exposed to someone - it is possible for anyone at all times to begin to develop the series? Is understanding of the series possible if and only if the set of conditions is fulfilled and situations when a person understands the series earlier - before all the supposedly required conditions are fulfilled - or later - after some additional conditions must have been fulfilled - are unimaginable? The last two questions should make the essentialist feel something is wrong with the conception of essence. His position is difficult to maintain not only because every proposed set of conditions may appear to be too weak but also, what is equally important, because every set may appear to be too strong. Understanding may be gained both on a higher and a lower level than the essentialist expects. Essentialism, as seen by Wittgenstein, consists /1/ in assuming that for a multiplicity of different objects which fall under a category, there exists an element common to them all as well as elements specific for every one of them, /2/ in crediting both such a common element and specific elements with exceptional value in guaranteeing the understanding of an object chosen from the multiplicity and /3/ in commanding quests after these credited with value elements. As for the notion of understanding itself mistakes of essentialism lie in treating the phrase "understanding of X" in analogy to a function with argument X instead of in analogy to a name with index X*2*. And in creating the convictions that the grasping of an element - the essence of understanding - may mean the understanding of understanding or that the momentary discovery may guarantee understanding lasting for an unlimited time. These mistakes have been noticed and internalized by the interpreters of Wittgenstein. But the most dangerous mistake, lying in the demand to discover an equivalent of understanding, is still being toyed with. The conviction remains that we can discover a factor independent of ourselves which is responsible for understanding, a factor which settles the question of understanding in every case with the required, definitive accuracy. As long as the conviction appeals to us we are prevented from noticing that the factor bringing the definitive accuracy is none other than the human factor. Standard interpretations underline that Wittgenstein's constructive considerations, in consequence of criticism of mental essence, circulate round two postulates. Firstly, the problem of "understanding of X" finds its settlement in the sphere of outer facts. Secondly, ordinary facts determine "understanding of X" and there is no such fact that would be the understanding of X in itself and, therefore, extraordinary. No matter where the postulates are rooted, whether in the Private Language Argument or elsewhere, they have an anti-essentialistic flavour. They immediately direct one towards Wittgenstein's famous thesis that the phrase "understanding of X" is often referentially empty and acquires its meaning only via transitive explanations that reveal its criterial links with some descriptive phrases*3*. Thus the route Wittgenstein takes inevitably leads from anti- essentialism to the delicate issue of criterial links between declarative sentences, or, to present the problem in the objective mode, between facts. Delicate, as criterial links are judged to be necessary - non-inductive - which may impel one to formulate some questions: in what sense are criterial links necessary? Are criterial links a priori? How does one discover criterial links if they are necessary a priori?*4* These questions are dangerous since they provoke us to cross the path of essentialism again or resign to conventionalism. Though we submit to the first of Wittgenstein's postulates and focus on "the outer" we remain convinced that criterial links should be discovered. We ask how to do it and stubbornly attempt to unravel this mystery or, disheartened, give up and accept a solution substitute, such as convention. For Wittgenstein, I believe, the idea of discovering criterial links between facts is absurd for the very special reason that criterial links do not occur just between facts but between facts transformed into institutions. It is a remnant of essentialism that we treat criterial links as being primary beyond our control saying that they are either natural, included in the world itself, or artificial, created by people themselves, but located in the private area of mind. The obsession to establish control over every domain that is beyond our control stimulates the need for discovery where discoveries are logically impossible: where the solution lies mainly in the description of the manner we create the linguistic institutions which are not conventions. It is not my intention to suggest that there is only one way of creating linguistic institutions. Nevertheless, there is one way which I would like to describe because it connects closely the issue of a criterion with that of understanding. To be more concrete: Wittgenstein's philosophy of understanding, I assume, has its roots in the idea that in order to ascribe understanding to someone it is necessary for him either to be taught, to be tested by someone else who takes the role of a teacher, or to teach, to be tested as a teacher himself. A person who is never able to be successfully taught something by someone or successfully to teach someone something understands nothing. If one maintains that understanding is contingent upon teaching*5* and successful explanation*6* one accepts cases of understanding which are completely unintelligible. If one identifies criteria merely with grammatical conventions*7* - results of agreement within communities - one is in no better position because the term "convention" is completely unsuitable for the explanation of some occurrences at the boundaries of language where agreement is not a result of discussion or collective decision. It is, therefore, worth trying to see in teaching a necessary instrument producing understanding exactly by transforming descriptions of facts into linguistic institutions, i. e., criteria. Unfortunately, the mistaken picture of teaching we have hinders the recognition of this message. According to Wittgenstein, we usually hold that teaching results in a mysterious mental entity that had not been previously included in it. Teaching, the activity producing understanding, seems to fill the gap between two systems of signs, the one presented by a teacher and the one with which a learner responds, and understanding emerges as the third entity which must, subsequently, be localized and examined. That is why we ask: what is understanding in itself? Wittgenstein denies that a third entity is unconditionally necessary or sufficient for ascribing understanding to someone. A teacher who presents examples for his learner has, in advance, prepared a vision of the reactions he intends to generate. In order to be valid the reactions cannot be unpredictable or inaccessible for the teacher. Every third entity which comes into existence during teaching is an accidental, secondary side effect. Teacher's presentations are, in this sense, the reasons why understanding occurs and all phenomena which additionally appear are, at most, the causes of understanding*8*. The reasons cannot be discovered: they are chosen, arranged by the teacher, given to the learner directly and their number is finite. The causes can be discovered: they are hidden and their number is potentially infinite. Mental phenomena are neither the reasons why understanding occurs nor the causes of understanding because they cannot be used by the teacher as his chosen tools or be revealed in the learner's responses. The teacher can never try to tell the learner what phenomenon should appear in his mind to make him understand, neither can the learner tell what phenomenon has appeared there to convince the teacher that he understands. If they tried to do just that they would be idiots. The distinction "reason - cause" (criterion - symptom) may suggest that the value of casual links is minimal. But this would be a simplification. It is hard to overestimate casual links because their appearance proves that the learner reaches autonomy. Discovering and revealing links which were not presented but are accepted by the teacher offers evidence that the process of teaching comes to an end. - The learner transcends what was given to him although all he discovers is still a part of the teacher's immanency. This teacher-learner relationship is replaced by a partnership or the previous roles are reversed when the learner transcends his teacher's immanency. What he presents then to his former teacher is reasons (criteria). What is said above allows me to point out an aspect of the problem of a priori. From the point of view of teaching, meant as a process either resulting in the learner's understanding or revealing the teacher's understanding, links between teacher's primary explanations are criterial and, what is more, are a priori for the learner in the sense that they are undiscovered, given to him directly. Symptomatic links - effects of discovery - presented to the teacher as responsive explanations by a learner are a posteriori. An interesting consequence of this conception is that criterial links can be a priori unilaterally. Needless to say, this interpretation of "a priori - a posteriori" I have drawn has little in common with the classical ones and may appear strange. To make it more acceptable I will place it now in a more familiar setting and propose some rationale for it*9*. The distinction of "a priori - a posteriori" emerges from the question: how can one know? which has at least two aspects. Firstly, what are the sources of knowledge? Secondly, what are the primary justifications of knowledge? Traditional empiricism, claiming that experience is the only source of concepts admits at the same time two kinds of propositional knowledge. Empirical knowledge, justifiable by experience, which is substantial - describes the world, and a priori knowledge, justifiable by reflection on relations between concepts, which is trivial - has nothing to do with the world*10*. Kant makes this division more complicated pointing out that experience has its pre-conditions. There are boundary concepts and connected with them propositional knowledge which is both a priori and substantial - synthetic - as it depicts the matrix which experience cannot transcend. This type of a priori knowledge, justifiable by the existence of pure reason, is in two ways necessary. Firstly, it consists of necessary truths - such that are impossible to be false. Secondly, it is the necessary background for empirical knowledge; knowledge consisting of contingent - possible to be false - truths. In Kant's system empirical knowledge appears to be contaminated by the alien a priori factor: without this a priori factor empirical knowledge would be impossible*11*. Contemporaries - for example S. Kripke - attack Kant's conviction that all necessary truths are universal. It is argued that some particular necessary truths are available not only by inference from universal truths, which is easy, but by reflection on an individual object. It is underlined that a priori knowledge is the knowledge which can be - but does not have to be - acquired without appeal to experience; whether a proposition known a priori can also be known empirically is irrelevant. There are no longer logical interdictions for the statement that a universal necessary truth can be known by inference from particular necessary truths. The thesis that necessary truths cannot be known empirically is treated as false in a strong sense: one can know directly from experience that a sentence is necessarily true. Thus, the meaning of the notion of a priori undergoes a process of far reaching liberalization and the connection between "a priori" and "necessary" is broken*12*. Finally, the notion of necessity is frontally attacked and reasons are given why no proposition can be immune from revision. For Quine necessary truths - truths impossible to be false - do not exist. There is no logical necessity. All knowledge is more or less empirical, contingency is a matter of degree, and there is at most natural necessity consisting entirely of what science produces*13*. It seems, therefore, that what has remained from the classical picture of the pair "a priori - a posteriori" is the thesis that contingent truths cannot be known a priori. They can be false. So, we cannot avoid empirical investigation to verify them. But this is not absolute. In this outline of the destinction "a priori - a posteriori" a striking thing is the permanent presence of the following assumptions: every declarative sentence is an epistemic problem in itself. Every sentence has to be classified according to its nature and justified since every one is believed to be defeasible. People build knowledge but the philosophical attitude towards it is unchangeably defensive: it seems impossible to find any place within the area of the empirical that would be really certain. Wittgenstein, I believe, is in opposition to this scheme. His grounds for depreciating this defensive attitude are rooted in questions which are as simple as these: is it sensible to require that every empirically verifiable sentence or conditional be always verified before it is asserted or used as a rule of reasoning? Is it true that a sentence or conditional discovered by experience never loses its empirical character? Wittgenstein's refutation makes the thesis that contingent truths cannot be known a priori collapse. This refutation is to a high degree determined by the logic of teaching. - If teaching both gives and reveals understanding it also must change the epistemic status of empirical sentences. It is worth mentioning that if we adopted the notions of an ideal teacher, someone who knows everything that can be known, and an ideal learner, someone who knows nothing at all, we would have to say that every set of empirical sentences the ideal learner is given when teaching begins includes sentences that are for him synthetic a priori, i. e., give the form for the learner's experience. The perspective of teaching, in other words, changes the classification of sentences. Although in a transient manner, some sentences that would be otherwise classified as empirical and contingent present themselves as a priori and necessary. The logic of teaching I ascribe to Wittgenstein is dominated by the distinction "criterion - symptom" (reason - cause) as well as by the principle of balance. As a result of the distinction "criterion - symptom" the problem of someone's understanding looks relatively simple as far as criteria of understanding are concerned. When we demand symptoms it may look infinitely complex. When we mistake symptoms for criteria and look for criteria among symptoms the problem of understanding is insolvable. The principle of balance says that explanations may be effective only if they are picked up from a language game, transformed into criteria and then put into the game again, i. e., are finally recognizable - now from the point of view of the language game as a whole - as examples of rules of the game or moves within the game. Thus, criteria appear to be result of making individual decisions within the game rather than the matter of the nature of explanations. In particular, criteria need not be the rules of a game that is taught but can belong to the category of moves. In connection with the distinction "criterion - symptom" Wittgenstein talks, especially in ON CERTAINTY, about the role of primitive attitudes and the point of a game. Teaching has a real subject - a language game - and a background - circumstances in which it takes place. A preliminary condition for teaching to come to an end is that both sides involved must focus on the subject while ignoring the background. This means that the subject must be unanimously regarded as worth teaching and learning. There are two different levels of teaching with their own distinct tasks. Firstly, the learner is taught rules of the game to obviate evidently nonsensical moves. Secondly, he is taught various strategies of reaching the point. Ineffectiveness in reaching the point in changeable circumstances is equally disqualifying as the breaking of rules. Inability to transcend teacher's examples, to make original moves in order to reach the point, is a manifestation of the nonsense of playing. At first sight the notion of the nonsense of playing is not interesting. But when we accordingly change the way of seeing it becomes so. It is enough to forget for a moment about the level of language games and ascend the higher level of language as a whole. Then the new perspective immediately makes us ask questions such as: what is the point of language? What is the strategy of realizing the point? If we additionally admit that the kernel of language is declarative sentences our provisional answer to the first question is: the point of language is the stating of truth. And to the second: the strategy is determined by the method of acquiring knowledge. - He who knows nothing also understands nothing. Epistemic reflection, Wittgenstein says, offers us an ideal of knowledge which we believe is worth accepting even if it seems to be unattainable. In the case of indirect knowledge our intuition is that we know that q if /1/ the connection between p and q is necessary and /2/ p is, to use a Kantian term, apodictically true. We think that knowledge demands an infallible scheme of inference together with a foundation made of apodictic truths. Apodictic truths fascinate us also for a different, though equally important reason: we are convinced that these truths stabilize language by guaranteeing the firmness of meaning for words. It is a philosophical truism that ordinary empirical sentences, which are contingent truths, are suitable for none of the purposes. What would be most suitable is sentences synthetic a priori. In ON CERTAINTY Wittgenstein confronts these imperatives with his conception of World Picture, the factor stabilizing language*14*. And by no means accidentally does he connect it with the theme of the nonsense of playing. A standard of nonsense we are accustomed to is a contradictory sentence; a sentence that cannot be true. The most perfect embodiment of sense, on the other hand, is for us an apodictic truth; a sentence that cannot be false. Wittgenstein is suspicious of the second part of this clichי. On what grounds are some sentences taken to be apodictic truths? Is there really a place for apodictic truths in the game of knowing? He suggests the more intently we look at a truth contemplating its degree of certainty the more blind we are to the importance of its informativeness. Then we are unable to notice that sentences which could be recognized as apodictic truths have already become epistemically useless, i. e., ceased to be information*15*. This means they no longer are genuine moves in epistemic games with declarative sentences. They pervade the games but only because they create a remote background of thinking*16*. In no game can they be doubted but in no game can they be stated. The price of being apodictic is triviality. Informative sentences, on the other hand, the genuine moves, never attain the status of apodictic truths. Expected though they are, stating them as truths is risky and the game of knowing does little more than rationalize the risk. A gradation of sentences according to certainty within the game is always between less and more certain ones. Unconditionally certain sentences are never qualified to be elements of the game of knowing if knowing means informing. Thus, an attempt at installing them there is a manifestation of the nonsense of playing. There are a few passages in ON CERTAINTY that provide an illustration for these theses. Some sentences of the form of empirical sentences, Wittgenstein says, are withdrawn from the circulation of language since they have been so evident that they have become methodological sentences, i. e., a part of the method of thinking*17*. Methodological sentences differ from sentences still circulating in ceasing to be proper elements of the expression "X knows that...". These sentences are certain to the degree that not being a part of knowledge any longer because of their triviality they are not even taught but rather, in Wittgenstein's words, "swallowed"*18*. The category of methodological sentence determines a boundary of language consisting of sentences which once had been inside language but then, as worn out, were thrown out to its periphery*19*. This is one of the ways in which formally empirical sentences reveal their multidimensional nature. Clearly, Wittgestein's claim is not that some empirical sentences are at the same time a priori - this would be a contradiction - but that some empirical sentences are able to lose their former status and become paragons of truthfulness. If so, if some sentences appear to determine a boundary of language just because they drop out of circulation, it is naturally worth asking how the other end of the issue of boundaries of language looks. In other words: how do declarative sentences enter into circulation? Are all these sentences thoroughly empirical i. e., always possible to be false? It is enough to turn back to what Wittgenstein writes about the notion of a criterion and ostensive definition to find out that there are not such sentences. At the other boundary of language there are pseudo-sentences such as "This is X" or "If this is X then this is Y"*20*. These pseudo-empirical forms introduce samples, patterns, criteria or, more generally, elements of the method of identifying and making comparisons. Consequently, their linguistic status is very special. - The method of identifying and making comparisons makes us think. Declarative sentences enter into circulation via the filter of the institution of teaching and learning. Those which are transformed into components of ostensive definitions and defining criteria lose instantly, though temporarily, their empirical character and function as synthetic a priori. They regain their empirical character immediately when they are used as genuine moves within epistemic language games, i. e., when the question about their epistemic status makes sense again. The often repeated thesis that understanding is for Wittgenstein in an inner relation to communication attains a new dimension by adding that there are two inseparable forms of communication: teaching language games and playing them. The qualitative difference between both forms is that teaching is reserved for giving - and acquiring - the methods of thinking while playing is the domain of applying the methods. The difference is maintained although sentences belonging to the method and the domain of usage can be formally the same sentences. To perceive different aspects of one and the same sentence we must look at it from various perspectives. No one singular perspective reveals all the aspects. When viewing the term "declarative sentence" from the isolated perspective of a language game the instability of the meaning of the term is obliterated. Namely, that a declarative sentence may be empirical when seen as a move within an epistemic game, synthetic a priori when seen as an instrument in teaching the game and even analytic when seen as a triviality thrown out at the margin of the game. At least three perspectives can be inferred from the anti- essentialistic motif in Wittgenstein's philosophy of understanding. Each of them reveals different features of declarative sentences. 1. The perspective of teaching epistemic language games - a result of the manner and circumstances in which a declarative sentence is used can be transient synthetic apriority of the sentence. A person cannot be described as having understanding of an object - or fact - which was unknown to him unless the primary interpretation of it is imposed on him by a teacher or the primary interpretation of it is imposed by him on a learner, i. e., unless the person takes part in the form of genuine communication which is called teaching. 2. The perspective of playing epistemic language games - aposteriority of a sentence is prejudged neither by the fact of using the senses to grasp its meaning nor by its ability to turn into a falsehood. These two criteria of aposteriority have little in common with Wittgenstein's logic of playing epistemic games. For Wittgenstein, aposteriority is scaled according to the contribution of the sentence to the exchange of information by means of language. If, for example, a sentence is so commonly known that the normal way of realizing its logical value is by using language instead of using the senses or its logical value is so easy to realize that it would be strange to speak about discovering it or the sentence is so trivial that its logical value is normally not called into question then one cannot say that the sentence is fully empirical. 3. The perspective of language as a whole - aposteriority is a temporary quality attributed to sentences inasmuch as they are genuine moves in epistemic games but separated from them as they become elements of World Picture. The question whether we must use the senses to discover the meaning of a sentence should be preceded by another question: is the effort justified? Sentences that are not expected to be discovered at all are, on the one hand, these which are given during teaching as bearers of the methods of thinking and, on the other hand, these about which people stop thinking because they are for them the ultimate patterns of certainty and triviality. Their truthfulness is not discussed because it is always too early or too late to discuss them. In summary, the conditions we have to meet to enter into language seem to determine entirely what understanding is. In fact, we do not enter by ourselves but are introduced into language and some autonomy is subsequently given to us. From then, however, the thread of the loss of understanding hangs over us permanently and the loss is a fact first and foremost when we begin to ramble through language misinterpreting its boundaries and not realizing its point. Wittgenstein shows the other side of understanding by attempts to draw the conditions we have to meet in order not to ramble through language. FOOTNOTES *1* G. P. Baker and P. M. S. Hacker, An Analytical Commentary on Wittgenstein's PHILOSOPHICAL INVESTIGATIONS /Oxford, 1958/, pp. 267 - 317 *2* L. Wittgenstein, The Blue and Brown Books /Oxford, 1958/, p. 21 *3* L. Wittgenstein /1958/, pp. 24 - 25 *4* P. M. S. Hacker, Inside and Illusion /Oxford, 1989/, pp. 310 - 322 *5* G. P. Baker and P. M. S. Hacker, Wittgenstein - Meaning and Understanding /Oxford, 1984/, p. 31 *6* G. P. Baker and P. M. S. Hacker /1984/, p. 38 *7* P. M. S. Hacker /1989/, p. 309 *8* L. Wittgenstein /1958/, pp. 14 - 15 *9* J. Dancy, Introduction to Contemporary Epistemology /Oxford, 1985/, pp. 212 - 225 *10* J. Dancy /1985/, pp. 212 - 214 *11* J. Dancy /1985/, pp. 214 - 218 *12* J. Dancy /1985/, pp. 219 - 221 *13* J. Dancy /1985/, pp. 222 - 225 *14* L. Wittgenstein, On Certainty /Oxford, 1969/, §§ 95, 162, 167 *15* L. Wittgenstein /1969/, §§ 460, 461, 553 *16* L. Wittgenstein /1969/, §§ 411, 105 *17* L. Wittgenstein /1969/, § 318 *18* L. Wittgenstein /1969/, § 143 *19* L. Wittgenstein /1969/, §§ 210, 211 *20* L. Wittgenstein /1969/, §§ 53, 57, 412