Phenomenology and Fallibility

Husserl Stud (2009) 25:1–14 DOI 10.1007/s10743-008-9053-3 Phenomenology and Fallibility Walter Hopp Published online: 17 February 2009
Springer Science+Business Media B.V. 2009 Abstract If Husserl is correct, phenomenological inquiry produces knowledge with an extremely high level of epistemic warrant or justification. However, there are several good reasons to think that we are highly fallible at carrying out phenomenological inquiries. It is extremely difficult to engage in phenomenological investigations, and there are very few substantive phenomenological claims that command a widespread consensus. In what follows, I introduce a distinction between method-fallibility and agent-fallibility, and use it to argue that the fact that we are fallible phenomenologists does not undermine Husserl’s claims concerning the epistemic value of phenomenological inquiry. I will also defend my account against both internalist and externalist objections. The discipline of phenomenology presents us with a bit of a puzzle. On the one hand, Husserl insists that phenomenological inquiry results in knowledge of a very extraordinary type. Perhaps Husserl does not, as Dagfinn Føllesdal has argued, hold that any knowledge is infallible, and given the deep skepticism regarding infallible knowledge that pervades contemporary philosophy, I will not pin such a thesis on Husserl. Nevertheless, it is clear that the knowledge we acquire through phenomenological reflection does, according to Husserl, possess a degree of warrant appreciably greater than we could hope to attain in the sphere of ordinary empirical objects. The objects of phenomenological inquiry, namely the essential features of types of conscious experience, are supposed to be given to us, and given in the most complete manner possible. Phenomenological cognition is, if not infallible, very nearly so, and much more nearly so than, say, your knowledge that you have a left hand. W. Hopp (&) Department of Philosophy, Boston University, Boston, USA e-mail: hopp@bu.edu 123 2 Husserl Stud (2009) 25:1–14 On the other hand, Husserl himself admits that phenomenological inquiry is exceedingly difficult, that the risk of error confronts us at every step, and that virtually every concrete result of philosophical substance is a hard-won achievement. This is something that we can verify for ourselves simply by doing phenomenology. Moreover, there exist widespread disagreements across a number of phenomenological subject matters, such as the nature of perception, consciousness, self-knowledge, and so forth. And so phenomenological cognition is supposed to be nearly infallible, and yet we are very fallible phenomenologists. Is such a position coherent? I believe it is. By distinguishing between method-fallibility and agent-fallibility in what follows, I will argue that the fact that we are fallible phenomenologists does not entail that the phenomenological method itself is unsuited to discover the essential features of lived conscious experiences. I will also argue that the fact that we are agent-fallible with respect to the phenomenological method is perfectly consistent with the claim that we possess knowledge with a very high degree of epistemic warrant when, on occasion, we carry out that method properly. 1 Phenomenological inquiry is often exceedingly difficult, and presents us with many opportunities to get things wrong. Husserl was well aware of this. In the Introduction to Volume II of the Logical Investigations, Husserl writes that phenomenology involves ‘‘reflection,’’ that is, ‘‘mak[ing]… acts themselves, and their immanent meaning-content, our objects’’ (Hua XIX/1, p. 14; 1970, p. 255), a mode of thinking he characterizes as ‘‘unnatural’’ (Hua XIX/1, p. 14; 1970, p. 254). As if engaging in an already unnatural mode of thinking were not difficult enough, he adds that we must ‘‘strive to arouse dispositions in ourselves which will keep our meanings unshakably the same, which will measure them sufficiently often against the mark set by reproducible intuitions…’’ (Hua XIX/1, p. 10; 1970, p. 252). And, in The Idea of Phenomenology, he indicates what sorts of acts the phenomenologist must carry out: Phenomenology carries out its clarifications in acts of seeing, determining, and distinguishing sense. It compares, it distinguishes, it connects, it places in relation, it divides into parts, it separates off moments. But it does all this in the act of pure seeing. It does not engage in theory or mathematical construction; that is, it offers no explanations in the sense of deductive theories. (Hua II, p. 58; 1999, p. 43) And, of course, in his more mature work Husserl outlines a rigorous method involving multiple reductions in order to get us to carry out the acts required to make transcendental consciousness our theme. Husserl gives us an idea of what we, as aspiring phenomenologists, are up against: It is not only that, prior to any method for determining matters within its field, a method is needed in order to bring, without exception, the field of affairs 123 Husserl Stud (2009) 25:1–14 3 pertaining to transcendentally pure consciousness within the regard which seizes upon it; it is not only that this requires a difficult turning of the regard from the natural data which continue to be objects of consciousness and are thus, as it were, interwoven with the data newly intended to, so that the danger of confusing the two sets of data is always threatening; but it is also that everything helpful to us in the case of the natural sphere of objects is lacking… (Hua III, p. 150; 1982, p. 148) Whatever else phenomenological inquiry is, then, it certainly involves a great deal more than staring at essences or waiting around for the relevant objects and states of affairs to impress themselves upon us. The circumstantial evidence that we are fallible phenomenologists is, if not overwhelming, at least very strong. Even if we grant, as I think we should, that what Descartes, Locke, Berkeley, Hume, Reid, Kant, Brentano, and legions of contemporary philosophers have done and are doing is not precisely phenomenology as Husserl envisions it, they are, at least at times, in the same general line of business. This is clear from the fact that they make claims that Husserl, and any other serious phenomenologist, must bother to refute or confirm, and often make them on the basis of an allegedly direct ‘‘inspection’’ of consciousness itself. Yet when one looks at the various claims about the structure of intentionality, the nature of perception, and so forth put forth by various philosophers, one can only be struck by the lack of any overarching consensus on virtually every significant issue. Indeed, it ought to strike us as extremely puzzling that claims concerning the broadly physical facts concerning, say, perception command a very impressive consensus, despite the fact that the relevant states of affairs that those claims are about do not feature in our lived experience, while there are precious few substantive philosophical issues concerning the phenomenological character of perception that most unbiased investigators can agree upon. This cries out for an explanation. To be sure, there are several explanations for this lack of consensus that have nothing to do with the epistemic credentials of phenomenological reflection as such. Disagreements can arise because of a certain blindness to the phenomena—and Husserl’s theory of givenness, as explained below, certainly allows for blindness of that sort. Disagreements can also arise for ideological reasons. If one is wedded firmly enough to a world-view in which consciousness, meanings, essences, and the rest cannot find a home, then one might very well deny their reality despite ‘‘seeing’’ them. The charge of ‘‘feigning anesthesia’’ is cast quite often, and no doubt sometimes hits the mark. And, of course, some disagreements might be apparent only. That our words have the same sense, that they aim at the same objects and are fulfilled on the same intuitive basis, can never be taken for granted in phenomenology. Still, I doubt that the controversies that plague phenomenological inquiry can all be explained under these or similar headings. That phenomenological inquiry is difficult to carry out, presents opportunities for significant, even massive, error, requires practice, and constitutes an unnatural mode of cognition is something that is easy to verify without relying on the authority of Husserl or the sociological facts about the phenomenological enterprise. Anyone who has attempted to practice it independently, or even to verify the claims of other phenomenologists—helping 123 4 Husserl Stud (2009) 25:1–14 himself, naturally, to the concepts, vocabulary, and distinctions already formulated and drawn by another—must be intimately familiar with just how monumental a task it is even to put oneself in the proper frame of mind to do it properly, and how so much remains vague, dubious, unclear, and uncertain even when one has done so. Husserl, for instance, seems to think that the visual field consists of a twodimensional manifold.1 But for the life of me, no matter how readily I can think that the manifold visually given to me is two-dimensional, I cannot help but see it as three-dimensional. One of us is surely mistaken, and it may very well be me—but still, a mistake is occurring somewhere. And deciding this issue pales in comparison with a question like ‘‘What features are essential to visual consciousness as such?’’ Are there three essential features? Three hundred? Do we even have the concepts that we need to designate the features we will find when we begin looking? And won’t finding those features be much more difficult if we lack those concepts? Often, without the guidance of a Husserl or a Heidegger or a Gurwitsch, one can only feel paralyzed in the face of such questions. And yet, we are assured, the objects of phenomenological inquiry are given, and given in a strict sense. The phenomenological reduction, Husserl explains, signifies the limitation to the sphere of pure self-givenness, to the sphere of what is not merely talked about and referred to; but also not to the sphere of what is perceived, but rather to what is given in exactly the same sense in which it is meant—and self-given in the strictest sense—in such a way that nothing that is meant fails to be given. In a word, it is a limitation to the sphere of pure evidence, ‘‘evidence’’ here understood in a strict sense that excludes ‘‘mediate evidence’’ and, above all, evidence in the loose sense. (Hua II, pp. 60–1; 1999, p. 45) For an object to be given in the strictest sense—given in precisely the way in which it is meant, and meant in just the way in which it is given—is, according to Husserl, the ideal of adequate evidence (see Hua XIX/2/1970, Sect. 38). The objects of ordinary perception do not permit themselves to be given in this way, since there is always more to them than what is strictly given intuitively in any perception of them. Experiences of this sort possess evidence in a looser sense (see Hua XIX/2/ 1970, Sect. 38 and Hua III/1982, Sect. 138). In the Cartesian Meditations, Husserl distinguishes adequacy from apodicticity or ‘‘absolute indubitibility’’ (Hua I, p. 55; 1977, p. 15). An apodictic evidence is not only certain evidence of its objects but is also ‘‘the absolute unimaginableness (inconceivability) of their non-being’’ (Hua I, p. 56; 1977, p. 16). Not all apodictic evidences are adequate—apodicticity ‘‘can occur even in evidences that are inadequate’’ (Hua I, p. 55; 1977, p. 15)—but Husserl does appear to hold that all adequate evidences are apodictic. There is simply no room for reasonable doubt or uncertainty once the ideal of adequacy has been attained. Nor, when some proposition has been adequately fulfilled in a corresponding intuition, is there any need to prove, through deductive or any other sort of reasoning, that it is true. This is why Husserl characterizes phenomenology as a science taking place ‘‘within the limits of mere immediate Intuition, a purely 1 See, for instance, Hua XVI, p. 117; 1997, p. 98. 123 Husserl Stud (2009) 25:1–14 5 ‘‘descriptive’’ eidectic science…’’ (Hua III, p. 153; 1982, p. 150). If this ideal is what the phenomenologist does achieve, then phenomenological cognition ought to be much less fallible than we know it to be. 2 How can this be? There are a couple of reasons why we, as individual thinkers, could be fallible phenomenologists. Either (a) we are rather poor at carrying out the mental acts prescribed by the phenomenological method, or (b) that method itself is unsuited to discover the essential features of lived experiences. To put it another way, the source of our fallibility could be that it is difficult, perhaps exceedingly so, for us to carry out the required intentional acts to have phenomenological data given to us in an evidential manner, or it could be that phenomenological data do not give themselves to us in an evidential manner even when we carry out the acts prescribed by the phenomenological method in the right way. When it comes to acquiring phenomenological knowledge—or any knowledge, for that matter—there are at least two distinct sorts of possible impediment: we could either be inept at wielding the tools required for the job, or the tools themselves might be defective. I will henceforth refer to these sources of fallibility as agent-fallibility and method-fallibility.2 A method M is method-fallible with respect to some subject matter S if carrying out the mental acts prescribed by M does not result in warranted beliefs about S. For instance, the activity of staring at one’s foot is method-fallible with respect to the subject matter of modal logic, since staring at one’s foot does not yield warranted beliefs about modal logic. It is, however, a good method for discovering certain features of one’s foot. A method that is not method-fallible with respect to S is warrant-generating or, to borrow a phrase of Alston’s, ‘‘epistemizing.’’ A method that is epistemizing with respect to some subject matter is one which, when carried out properly, is capable of generating beliefs about that subject matter which possess a high degree of epistemic warrant. A thinker T is agent-fallible with respect to some method M if T is incapable or barely capable of carrying out the mental acts prescribed by M. So, for instance, a beginning student in mathematics is agent-fallible with respect to the methods of adding and subtracting numbers. The opposite of this is agent-reliability: a thinker is agent-reliable with respect to some method if it lies within the scope of that thinker’s power to carry out the cognitive acts prescribed by that method most of the time he attempts to do so. There are, thankfully, some methods of acquiring beliefs which are both epistemizing with respect to some subject matter and with respect to which we are reliable. For example, forming beliefs about ordinary middle-sized objects on the basis of perceptual experience is epistemizing, and is nearly as easy as being awake. There are also methods with respect to which we are all too reliable, but which are themselves not epistemizing with respect to any subject matter of 2 Thanks to David Jennings for suggesting the phrase ‘‘agent-fallibility.’’ David Roochnik has pointed out that this is somewhat misleading, since to be fallible at something at least implies the possibility of success, whereas the limiting case of agent-fallibility, as I am construing it, is sheer incompetence. Provided we bear this in mind, however, I think the term will not mislead. 123 6 Husserl Stud (2009) 25:1–14 worth. These include, among other things, hasty generalizations and self-serving rationalizations, at which we are all more or less expert. 3 The obvious payoff of this distinction is that we can, at least hopefully, account for our fallibility as phenomenologists without damning the phenomenological method itself. If we are fallible phenomenologists, and if phenomenology is epistemizing with respect to the essential features of lived conscious experience, then we had better be agent-fallible with respect to that method. For suppose that the reason for our fallibility were that the phenomenological method is method-fallible with respect to the essential features of lived conscious experiences. Such a claim would be fatal to phenomenology itself. If it were true, then even someone as agent-reliable as God could not, by employing that method, arrive at warranted beliefs about the essential features of consciousness. Daniel Dennett, for one, suggests that the ‘‘embarrassing fact of controversy and contradiction’’ in phenomenological contexts is explained by the fact that ‘‘we are fooling ourselves about the high reliability of introspection…’’ (Dennett 1991, p. 67). Dennett, as many have persuasively argued, is simply mistaken in identifying the phenomenological method with introspection, and, given some plausible arguments against the reliability of introspection, correcting him on this point is of enormous consequence.3 However, this still leaves part of his argument untouched: after all, whatever the phenomenological method turns out to be, those who attempt to employ it often disagree with one another, and with different stages of themselves. The problem with Dennett’s argument is not just that it contains a false premise, but that the conclusion does not follow from the premises. What follows from the aforementioned ‘‘embarrassing fact’’ is not that the phenomenological method (or introspection) is method-fallible, but either that it is method-fallible with respect to the essential features of consciousness, or that we are agent-fallible with respect to it. The challenge now, however, is to explain how we could be agent-fallible with respect to the phenomenological method. It would be one thing to make this distinction with respect to mathematical reasoning or determining whether one can mate in five moves, since such procedures involve articulated steps. Mathematical reasoning is epistemizing with respect to determining truths about numbers, and yet many of us are agent-fallible with respect to it. But this permits of an easy explanation: mathematical reasoning proceeds in steps, and errors can occur along the way. Husserl, however, insists that phenomenology does not proceed deductively or inferentially; it confines itself to what can be given, and given in a strict sense. Phenomenology takes place ‘‘within the limits of mere immediate Intuition…’’ (Hua III, p. 153; 1982, p. 150). Of course, Husserl does employ arguments from time to time, but their point is to lead us to see the matters in question, not to venture beyond what is given. 3 On the claim that phenomenological reflection is not introspection, see Thomasson (2005), Zahavi (2007), and Drummond (2007). 123 Husserl Stud (2009) 25:1–14 7 [D]eductive theorizings are excluded from phenomenology. Mediate inferences are not exactly denied to it; but, since all its cognitions ought to be descriptive, purely befitting the immanental sphere, inferences, non-intuitive modes of procedure of any kind, only have the methodic function of leading us to the matters in question upon which a subsequent direct seeing of essences must make given. (Hua III, p. 173; 1982, p. 169) But how, if the relevant objects and states of affairs are given to us, is there any room for agent-fallibility? If the traditional ‘‘mythic’’ conception of givenness were the right one, this problem would be intractable. On that conception, something’s being given is a purely passive phenomenon, a simple matter of being awake and in the thing’s presence. Acquiring knowledge in this way does not, according to the myth, presuppose any prior cognitive achievements, such as prior knowledge or the acquisition of concepts, or any specific activity on the part of the knower. This conception has been badly beaten up in the course of its career, and I think that, at the end of the day, it is best to retire it. This, however, is not Husserl’s conception of givenness. Whatever Husserl’s transcendental idealism says about the ontological status of the objects of knowledge and perception, it at least says this much: corresponding to any type of object whatsoever there are types of intentional acts in which that object could be brought to givenness or ‘‘constituted.’’4 These intentional acts, moreover, intend their objects, in the way that they do, in virtue of their own immanent parts and features. If an object is given to me in a certain way, that is not only because I am conscious and it is there, but because my act of perceiving it has the appropriate internal parts and features to qualify as a perceiving of it. Just as grasping an object with one’s hand requires one’s hand to be configured in suchand-such a way, so mentally grasping an object requires one’s mind to be configured in such-and-such a way, and one of the major tasks of phenomenology is to describe just how the mind must be configured if it is to lay hold of the objects that it does. As opposed to every theory of consciousness that treats it as a sort of blank slate waiting for objects to impress themselves on it, or a mirror whose job is solely to reflect whatever is present to it, Husserl regards every conscious act as having an intrinsic structure, and owing much of its intentional character to that structure. Consciousness and its objects are, in a certain sense, made for one another in much the way that a key and a lock are—which is not, of course, to say that one makes the other.5 This aspect of Husserl’s account of intentionality is idealistic in the sense 4 ‘‘[S]eeing consciousness… is just acts of thought formed in certain ways, and things, which are not acts of thought, are nonetheless constituted in them, come to givenness in them…’’ (Hua II, p. 72; 1999, p. 52). As this passage suggests, to constitute an object in consciousness is to carry out those acts in which it is brought to givenness, which need not be a process of making or constructing it. 5 See Hua III, p. 340; 1982, p. 333: ‘‘To every region and category of alleged objects there corresponds phenomenologically not only a fundamental sort of sense… but also a fundamental type of originarily presentive consciousness of such senses and, belonging to it, a fundamental type of originary evidence which is essentially motivated by originary givenness of such a character.’’ David Woodruff Smith holds that everything in the world has, in addition to a form and a substrate, an ‘‘appearance,’’ which ‘‘is how it is known or apprehended’’ (Smith 2004, p. 17). An appearance, then, is part of a formal ontological structure ‘‘that applies to any entity in our kind of world,’’ and is, therefore, what the Scholastics would call a ‘‘transcendental’’ (p. 28). 123 8 Husserl Stud (2009) 25:1–14 that it recognizes that there are quite definite noetic and noematic conditions for the possibility of intentionality and knowledge—that neither intentionality nor knowledge comes ‘‘ready-made.’’ It does not, however, entail that the existence or nature of stars, numbers, or God depends upon those noetic conditions, any more than a good theory of grasping entails that the properties of branches and rocks depend upon the configurations of human hands. What we ‘‘make’’ in constituting objects are not the objects themselves, but the intentions in virtue of which we can be conscious of them.6 Furthermore, in many, perhaps the vast majority, of cases such constituting acts are internally structured, composed of founded, founding, and mutually founding act-moments and parts, and essentially related to other acts. ‘‘The great theme of transcendental philosophy,’’ Husserl writes, ‘‘is consciousness in general as a storied structure of constitutive accomplishments in which ever new objectivities… are constituted in ever new levels or layers…’’ (Hua XI, p. 218; 2001, p. 269). Note here that objects are constituted, i.e., brought to givenness, in stages or levels, and this is true even in the phenomenological sphere. The method of clear seizing upon phenomenological essences themselves ‘‘requires proceeding step by step’’ (Hua III, p. 159; 1982, p. 156). On this view, then, givenness is not, to borrow a phrase from Blackburn, the ‘‘bare unvarnished presentation of fact to a purely receptive mind’’ but can be, and often must be, mediated by other acts (Blackburn 1984, p. 233). Some objects, such as universals and states of affairs, can only be given in founded acts. These objects do not just announce themselves to any mind of any type; they will only give themselves when the consciousness in question has the appropriate shape and form, and has carried out the necessary labor, to receive them. However, the sort of ‘‘mediation’’ envisioned here is ontological, not epistemic.7 Acts of givenness are non-inferential knowings, even though many are ontologically dependent, dependent for their whatness and their existence, on other acts. In light of Husserl’s account of givenness, the fact that we are fallible phenomenologists can be explained without impugning the phenomenological method itself. The fact that acts of phenomenological reflection might require work, and be founded in complex ways on other acts—including those acts in which we effect the various reductions—means that there are an awful lot of ways one can go wrong when carrying them out. We are agent-fallible with respect to the phenomenological method because it is difficult, for creatures like ourselves, to do each of the following: (a) perform the phenomenological reductions in order to prevent ourselves from ‘‘knowing’’ about our subject matter in advance of our inquiries; (b) fix our attention on our acts of consciousness in addition to their objects; (c) formulate concepts, carried out in the appropriate meaning-intentions, that uniquely pick out the features and properties of those acts; (d) fix those concepts, with just those meaning-fulfillments or fulfilling conditions, terminologically; and (e) perform eidetic abstraction and variation on the properties that we have managed to fix intuitively and conceptually. Our fallibility as phenomenologists might be 6 For an excellent discussion of Husserl’s realism, see Willard (2002). 7 See, for instance, Alston (1989). 123 Husserl Stud (2009) 25:1–14 9 due not to the fact that the phenomenological method is the wrong tool for the job, but to the difficulty creatures like us have wielding it. 4 This might be a satisfactory account of why we are fallible phenomenologists. It is, at any rate, surely part of the story. But explaining our fallibility is only half of the problem. The other half is to explain how, given this account, we can acquire phenomenological knowledge. For if Husserl is right, correctly using the phenomenological method to uncover the essential features of consciousness does result in knowledge. Surely he thought that he himself had acquired knowledge by these means, and so, I reckon, do many of us. Therefore, when we do employ that method successfully, we will acquire knowledge; the relevant objects and states of affairs will be given to us precisely as they are meant. But how, given this account, is that possible? A certain type of internalist about epistemic justification would insist that we cannot have it both ways. I will follow Laurence Bonjour in supposing that a theory of justification is internalist ‘‘if and only if it requires that all of the factors needed for a belief to be epistemically justified for a given person be cognitively accessible to that person, internal to his cognitive perspective; and externalist, if it allows that at least some of the justifying factors need not be thus accessible’’ (Bonjour 1992, p. 132). The internalist challenge I have in mind goes something like this: if we are, for the most part, agent-fallible with respect to the phenomenological method, then whatever beliefs we form when we attempt to employ that method will have a fairly low epistemic status. For, in order to form a warranted belief that p on the basis of some method M, I must also justifiably believe that (a) M is a reliable method, and (b) I have implemented M correctly, since those truths are, trivially, among the factors required for my belief that p to be epistemically justified.8 For instance, in carrying out a complex calculation, one’s warrant for believing the conclusion depends not just on the bare fact that one has carried it out properly. One must also justifiably believe that one has carried it out properly. After all, one’s justification for believing the conclusion can be defeated by the information that one has made a mistake somewhere along the line. But given that we are agent-fallible with respect to the phenomenological method, we will not have much of a justification for believing that we have carried it out correctly. And if one’s beliefs that one has carried out a putatively warrant-conferring method do not themselves have much in the way of warrant, then one’s beliefs in whatever conclusion one reaches by means of that method will not have much warrant either. One response to this objection is to say that the working phenomenologist, who practices the phenomenological method and thereby acquires a sort of expertise, will no longer be agent-fallible with respect to that method and so will satisfy the 8 This is meant to resemble Lawrence Bonjour’s (former) insistence that in order for any empirical belief B to be justified, one must justifiably believe that (a) B has feature / and (b) beliefs having feature / are highly likely to be true (see Bonjour 1985, p. 31). Bonjour has, of course, since abandoned his coherentist, but not his internalist, ways (see Bonjour 1999). 123 10 Husserl Stud (2009) 25:1–14 internalist constraints on warranted belief. Surely becoming less agent-fallible with respect to the phenomenological method is a demand placed on any aspiring phenomenologist, and I believe it is one that can be achieved. But this response misses a crucial philosophical consequence of the objection above: if the objection is sound, then Husserl has a massively flawed theory of givenness. Either (a) objects cannot be given by means of ontologically structured and articulated acts in the way Husserl supposes, or (b) they can, but givenness does not always have the epistemic status that Husserl says it does; the epistemic status of any act of fulfillment that is founded on other acts will be epistemically founded on one’s warranted beliefs that the method whereby one achieved that fulfillment is reliable, and that one has carried out the method properly. This response, then, saves Husserl’s view at one point only to have it devastated at another. If the version of internalism considered above is right, then Husserl’s theory of givenness cannot be. Fortunately, there are good reasons not to endorse an internalism of Bonjour’s variety; the internalist claim that all of the factors that contribute to the justification of a belief must be cognitively accessible to a believer is simply too strong to be credible, as I have argued elsewhere (Hopp 2008). For among the things that help explain why a given belief is epistemically justified or warranted is that it satisfies the antecedent of some true epistemic principle; indeed, that is constitutive of what it is to be an epistemically warranted belief. So on an internalist view, in order for my belief B to be warranted, two other things must be ‘‘cognitively accessible’’ to me: (a) there is some true epistemic principle P that specifies sufficient conditions for a belief to be warranted, and (b) B satisfies the antecedent of P. But this seems wrong for a number of reasons. First, there are many creatures, including those of the human variety, that seem to be knowers but also lack the concepts required to formulate epistemic principles or claims about their own psychological states. For some creatures, such as children, this might be a temporary cognitive handicap. For others, such as squids, it might be constitutive of their condition. Moreover, we need not even suppose that any such creatures actually exist. It is enough that such a consciousness is possible. And, it seems, such a consciousness is possible according to Husserl. For Husserl, consciousness and evidence are necessarily linked with one another. In the ‘‘extremely broad sense,’’ evidence is ‘‘an ‘experiencing’ of something that is, and is thus.’’ It ‘‘includes all experiencing in the usual and narrow sense’’ (Hua I, p. 52; 1977, p. 12). Evidence, construed broadly or loosely, does not require the actual attainment of anything like apodicticity or adequacy: my experience of the table before me is an inadequate experience of a table, not an adequate experience of something else. The failure to recognize the lower, assertoric grades of evidence that characterize the vast majority of conscious acts makes the accomplishments of consciousness a complete mystery. To ‘‘reduce evidence to an insight that is apodictic,’’ Husserl says, ‘‘is to bar oneself from an understanding of any scientific production’’ (Hua XVII, p. 169; 1969, p. 161). The important point here is that intentionality and evidence, and with them the having of warrant and possibly knowledge, are inextricably bound together. ‘‘Thus evidence is a universal mode of intentionality, related to the whole life of consciousness. Thanks to evidence, the life of consciousness has an all-pervasive teleological structure, a pointedness toward ‘reason’ and even a pervasive tendency 123 Husserl Stud (2009) 25:1–14 11 toward it…’’ (Hua XVII, p. 169; 1969, p. 160). All intentional processes whatsoever, then, must be understood as taking place within a ‘‘space of reasons.’’ If intentionality and the having of evidence are essentially related to one another, then the having of evidence, and with it warrant and knowledge, cannot be dependent upon either (a) higher-order acts directed upon first-order intentional states or (b) having warranted beliefs about epistemic principles. The reason is that first-order intentional acts do not depend, either for their existence or for their whatness, on reflective acts directed upon them, whether of the psychological or phenomenological variety. Their identity is, rather, presupposed by such acts. Nor do they depend upon one’s having any thoughts about epistemic principles. One can perceive, say, a table and thereby have (assertoric) evidence, without reflecting on that act or having even a passing thought about any epistemic principle. Indeed, that can occur without one’s even having the capacity to reflect on the act or to consider any epistemic principles. Because intentionality does not depend upon reflection or believing epistemic principles, having evidence does not either. If I am right about this, then there ought to be evidence that Husserl rejects internalism of this variety. And such evidence does exist. Consider his ‘‘Principle of All Principles,’’ which reads: ‘‘every originary presentive intuition is a legitimizing source of cognition… everything originarily… offered us in ‘intuition’ is to be accepted simply as what it is presented as being…’’ (Hua III, p. 52; 1982, p. 44). That is, if an object is exhibited intuitively to a subject S as S means it, then S is warranted in taking that object to exist and possess those properties that it intuitively appears to possess. What this principle does not require is that one must warrantedly believe the Principle of All Principles itself, or warrantedly believe that one’s own psychological states fall under it. The Principle is ‘‘not a principle I have to apply in order to gain knowledge; I need only fall under it’’ (Van Cleve 1979, p. 70). And this, I think, is just what we find phenomenologically. Suppose I come to believe that my desk is messy in virtue of seeing that it is messy. When I reflect on my belief, I discover that it has a certain epistemic status, and that it had that status prior to and independently of any reflection on it. My reflection does not endow it with that status. ‘‘All talk of knowledge,’’ remarks Husserl, ‘‘refers to a relationship between acts of thought and fulfilling intuitions’’ (Hua XIX/2, p. 735; 1970, p. 837), not to a relationship between acts of thought and acts directed upon those acts of thought. ‘‘If I call this intuited object a ‘watch,’ I complete, in naming it, an act of thought and knowledge, but I know the watch, and not my knowledge’’ (Hua XIX/2, p. 735; 1970, p. 837). 5 So much, then, for the internalist objection. Now, however, let us consider another objection, this time from the quarter of externalism. One of the most popular externalist theories of knowledge and/or justification, since the 1979 publication of Alvin Goldman’s seminal paper ‘‘What Is Justified Belief?,’’ is reliabilism. There are many varieties of reliabilism, but every version requires that in order for a belief to be justified, it must be brought about by a process that reliably produces true 123 12 Husserl Stud (2009) 25:1–14 beliefs. Unlike internalism, process reliabilism does not require that one know or warrantedly believe that a given belief is brought about by a reliable process in order for it to be warranted. It need only in fact be brought about by such process. Reliabilism, thus construed, is no threat to the position I am advocating here, since forming beliefs on the basis phenomenological reflection might very well be an extremely reliable process, even if we are rather unreliable when it comes to carrying out acts of phenomenological reflection. However, John Greco has advocated another version of reliabilism, agent-reliabilism, according to which ‘‘A belief p has positive epistemic status for a person S just in case S’s believing p results from stable and reliable dispositions that make up S’s cognitive character’’ (Greco 1999, pp. 287–288). This is a version of virtue epistemology. Just as virtue ethics makes the rightness of an action parasitic upon the character of the agent performing it, so virtue epistemology grounds the epistemic warrant of a particular belief in the ‘‘intellectual character of a cognizer’’ (Greco 1999, p. 287). However, this can give rise to an argument against the position adopted here. For suppose that carrying out the phenomenological method properly is not part of a stable set of intellectual dispositions on our part—suppose, that is, that we are agent-fallible with respect to it. Then, on at least one understanding of agent-reliabilism, the beliefs we do form when we do carry out that method properly will not have warrant, since those beliefs are not grounded in stable and reliable dispositions on the part of the agent.9 What motivates this view is the problem of strange and fleeting processes, whose possibility, according to Greco, shows that process reliabilism fails to specify a sufficient condition for warrant or knowledge. Here is one of Greco’s examples: Suppose that Rene devises a system for winning roulette that is based on the Gambler’s Fallacy, according to which a number is more likely to come up if it has not come up for a long string of spins. Rene has a helpful demon, however, who makes sure that whatever number Rene believes will come up will come up. Rene’s belief-forming process is reliable, but Rene does not have any warrant for his beliefs about which numbers will come up. The way to deal with this ‘‘strange and fleeting’’ process is to insist that one’s warranted beliefs arise from stable reliable dispositions. As Greco explains, ‘‘the cognitive faculties and habits of a believer are neither strange nor fleeting. They are not strange because they make up the person’s intellectual character… They are not fleeting because faculties and habits by definition are stable dispositions…’’ (Greco 1999, p. 287). I confess I am completely in the dark as to how this is supposed to deal with the Rene example above. For the problem with Rene’s method of forming beliefs is not 9 There is another understanding of agent-reliabilism according to which it poses no threat to the position here. Suppose that I am unreliable when it comes to performing acts of phenomenological reflection, but I habitually base beliefs on acts of phenomenological reflection on those occasions when I carry it out properly. Then, when I come to believe that p on the basis of properly carrying out an act of phenomenological reflection, we can either say (a) I am not justified in believing that p, because performing acts of phenomenological reflection is not part of my cognitive character, or (b) I am justified, because basing beliefs on phenomenological reflections is part of my character. If we opt for (b), then agent-reliabilism poses no problem for this position. However, it can also be construed along the lines of (a), in which case it does pose a threat. The problem here is that the definition of agent-reliabilism—in addition to entailing that we believe beliefs—is ambiguous between having a disposition to carry out acts of a certain type and having a disposition to believe on the basis of those acts when they are carried out. 123 Husserl Stud (2009) 25:1–14 13 that it was ‘‘strange and fleeting,’’ but that it was based on a fallacy. The alleged strangeness and fleetingness of Rene’s belief-forming processes have precisely nothing to do with why he lacks knowledge, and removing the alleged strangeness and fleetingness does not make him any more fit epistemically. It’s easy to imagine that Rene’s dispositions to reason that way are part of his settled intellectual character—and therefore neither strange, according to Greco’s rather strange characterization of strangeness—nor fleeting. But that, if anything, would make him worse off, not better. An agent who reliably bases his beliefs on fallacies is in worse epistemic shape than an agent for whom that is a strange and fleeting malady. Rene would not become a knower if he were routinely to reason according to a fallacy and, thanks to a demon, get things right. Rather, the demon would make him worse off epistemically (thought not financially) by giving him confidence in a bogus method of acquiring knowledge. And so, reliable processes, even when wedded to stable dispositions on the part of the agent, are not sufficient for warrant. More importantly, for our purposes, the reliability of an agent, his dispositions to form beliefs in a certain way, is not a necessary condition of his having evidence in a given situation. What matters is whether (a) the object intended in a person’s act is given to him in an evidential way, or, when the object is not given, (b) whether his belief is suitably grounded in such evidential presentations. That is, to have an object presented as one intends it is sufficient for warrant, even if one is unreliable when it comes to carrying out the acts in virtue of which such an object can come to givenness. For instance, I am agent-fallible when it comes to forming beliefs about the hidden images found in the once-popular Magic Eye stereogram images, because I am agent-fallible with respect to the method of having the images leap out at me. But when, on rare occasions, the hidden image does leap out at me, my beliefs about the content of that image have impeccable epistemic credentials. The fact that I cannot see such images most of the time, that I am quite unreliable when it comes to forming beliefs about their content, is irrelevant to the epistemic credentials of my beliefs when I do manage to see them. To wrap up, then: by distinguishing agent- and method-fallibility, I have attempted to provide the beginnings of a non-exhaustive explanation of how it is that phenomenological insights can have a very high degree of warrant, even though we are quite fallible phenomenologists. I have also attempted to explain how our beliefs can have a very high degree of warrant even when we are agent-fallible with respect to the method whereby we arrived at those beliefs. This is, of course, only a beginning: I have by no means established that the method of phenomenological reflection is in fact a good way of acquiring knowledge of the essential features of conscious lived experiences; I have only established that its being so is compatible with the facts that phenomenology is arduous and that the discipline itself is fraught with controversy and disagreement. Acknowledgments Thanks to David Jennings, Lynn Niizawa, and Irina Meketa for their extensive and helpful comments on an earlier draft of this paper. And thanks most of all to George Heffernan, whose meticulous and insightful comments have greatly enhanced my understanding of Husserl’s account of evidence. 123 14 Husserl Stud (2009) 25:1–14 References Alston, W. (1989). What’s wrong with immediate knowledge? In W. Alston (Ed.), Epistemic justification: Essays in the theory of knowledge. Ithaca: Cornell University Press. Blackburn, S. (1984). Spreading the word. Oxford: Clarendon Press. Bonjour, L. (1985). The structure of empirical knowledge. Cambridge: Harvard University Press. Bonjour, L. (1992). Externalism/Internalism. In J. Dancy & E. Sosa (Eds.), A companion to epistemology (pp. 132–136). Oxford: Blackwell. Bonjour, L. (1999). Foundationalism and the external world. Philosophical Perspectives, 13, 229–249. Dennett, D. (1991). Consciousness explained. Boston: Little, Brown and Company. Drummond, J. (2007). Phenomenology: Neither auto- nor hetero- be. Phenomenology and the Cognitive Sciences, 6, 57–74. Føllesdal, D. (1988). Husserl on evidence and justification. In R. Sokolowski (Ed.), Edmund Husserl and the phenomenological tradition (pp. 107–129). Washington: The Catholic University of America Press. Goldman, A. (1979). What is justified belief? In G. S. Pappas (Ed.), Justification and knowledge (pp. 1–23). Dordrecht: D. Reidel. Greco, J. (1999). Agent reliabilism. Philosophical Perspectives, 13, 273–296. Hopp, W. (2008). Husserl, phenomenology, and foundationalism. Inquiry, 51, 194–216. Husserl, E. (1963). Cartesianische Meditationen und Pariser Vorträge. Husserliana S. Strasser (Ed.). Den Haag: Martinus Nijhoff. Cartesian meditations (D. Cairns, Trans.). The Hague: Martinus Nijhoff, 1977. Husserl, E. (1966). Analysen zur passiven Synthesis. Aus Vorlesungs- und Forschungsmanuskripten, 1918–1926. Husserliana XI. M. Fleisher (Ed.). Den Haag: Martinus Nijhoff. Analyses concerning passive and active synthesis (A. J. Steinbock, Trans.). Boston: Kluwer Academic Publishers, 2001. Husserl, E. (1973a). Die Idee der Phänomenologie. Husserliana II. W. Biemel (Ed.). Den Haag: Martinus Nijhoff. The idea of phenomenology (L. Hardy, Trans.). Boston: Kluwer Academic Publishers, 1999. Husserl, E. (1973b). Ding und Raum. Husserliana XVI. U. Claesges (Ed.). Den Haag: Martinus Nijhoff. Thing and space (R. Rojcewicz, Trans.). Boston: Kluwer Academic Publishers, 1997. Husserl, E. (1974). Formale und transzendentale Logik. Husserliana XVII. P. Janssen (Ed.). Den Haag: Martinus Nijhoff. Formal and transcendental logic (D. Cairns, Trans.). The Hague: Martinus Nijhoff, 1969. Husserl, E. (1976). Ideen zu einer reinen Phänomenologie und phänomenologischen Philosophie. Erstes Buch: Allgemeine Einfuhrung in die reine Phänomenologie. Husserliana III. K. Schuhman (Ed.). Den Haag: Martinus Nijhoff. Ideas pertaining to a pure phenomenology and to a phenomenological philosophy. First book: General introduction to a pure phenomenology (F. Kersten, Trans.). The Hague: Martinus Nijhoff, 1982. Husserl, E. (1984). Logische Untersuchungen. Zweiter Band: Untersuchungen zur Phänomenologie und Theorie der Erkenntnis. Husserliana XIX U. Panzer (Ed.). The Hague: Martinus Nijhoff. Logical investigations (J. N. Findlay, Trans.). London: Routledge & Kegan Paul, 1970. Smith, D. W. (2004). Mind world. Cambridge, MA: Cambridge University Press. Thomasson, A. L. (2005). First-personal knowledge in phenomenology. In D. W. Smith & A. L. Thomasson (Eds.), Phenomenology and the philosophy of mind (pp. 115–139). Oxford: Clarendon Press. Van Cleve, J. (1979). Foundationalism, epistemic principles, and the Cartesian circle. The Philosophical Review, 88, 55–91. Willard, D. (2002). The world well won: Husserl’s epistemic realism one hundred years later. In D. Zahavi & F. Stjernfelt (Eds.), One hundred years of phenomenology (pp. 69–78). Boston: Kluwer Academic Publishers. Zahavi, D. (2007). Killing the straw man: Dennett and phenomenology. Phenomenology and the Cognitive Sciences, 6, 21–43. 123