I compile my text as I presented last Monday at the Value of Understanding Colloquium in Bahia. The idea hinges on whether perception can be measured by values that are not swamped by truth. To be sure, I use the scheme of a Zagzebskian Disjunctivist Externalism as a model: if Zagzebski's intellectual virtues are not swamped, I submit neither do perceptual virtues. If it is so, there is a possible disjunctivist response to the primary value problem. I think (pace Pritchard) this is an externalist response. Here is the text:
Epistemic externalism tends to make knowledge less dependent on the wonderings of a (human) knowing subject and more as an something attained in collaboration with the environment of the knower. Knowledge is less of a matter of inner cogitations and more an issue of interaction with what is to be known and its surroundings. In order to portray knowledge like that, knowledge-bearers such as beliefs (or opinions, convictions) have to be conceived as less than self-conscious states. However, as truth is normally conceived as being an external contributor to knowledge, the main focus of epistemic externalism has been further clauses that would distinguish knowledge from mere true belief. This means to discuss doxastic justification – or warrant, or truth-conducive procedures – and do it in a way that is not committed to the thesis that the knower has to be able to discriminate genuine cases of justification from deceiving appearances. In other words, in these provinces, the externalist endeavors therefore to show how there could be justification-contributors of which the subject of knowledge is not even potentially aware. One’s experience can be the same as somebody else and she can be more epistemically justified for reasons that are beyond her discrimination capacities.
In askeptical periods – those when, according to Linda Zagzebski’s (2001) terminology, the treat of skepticism is seen as dissolved or somehow innocuous – to investigate knowledge is no mainly focused on proving its possibility but rather on relating it to other notions and placing it among other phenomena. If the skeptic takes for granted that knowledge (or justification) is invaluable and wonders how it can be attained, in askeptical periods we are more likely to wonder why after all knowledge is to be sought as a goal, no matter if it is attainable. Epistemic externalism understands knowledge as firmly placed in the world around the knower rather than taking the knower to be to a large measure acting (or judging) from afar. Externalism is also akin to the idea that knowledge is not merely human, or at least not in principle dependent on anything specifically human. If knowledge has a value, it has to be somehow a value that is not tied to specific human faculties nor associated to distinctively human purposes. The value of knowledge would have to be construed not only out of something only humans can appreciate.
The value problem for externalists
In fact, the value problem has become a major problem for externalists recently. The general problem – or rather the class of problems – is old and familiar. In Meno, Plato questions whether knowledge is itself more invaluable than mere true belief. After all, in most cases, if one’s beliefs are true – about how to cook a mole or how to go back to one’s hotel in a Southern Hemisphere town – it is for most purposes enough. It is clear that it is not enough to come up with a property P that is itself invaluable and distinguishes knowledge from true belief for P can add no value to the mix. One has rather to consider how P would interact with a true belief. Kvanvig (2003: 45) explains why:
[…] consider some simple analogies. If we have a piece of art that is beautiful, its aesthetic value is not enhanced by having as well the property of being likely to be beautiful. For being likely to be beautiful is a valuable property because of its relationship to being beautiful itself. Once beauty is assumed to be present, the property of being likely to be beautiful ceases to contribute […] Take anything that you care about: happiness, money, drugs, sports cars, and so on. Then consider two lists about such things, the first list telling you where to obtain such things and the second list telling you where you are likely to obtain such things. Now compose a third list, which is the intersection of the first two lists. It tells you of ways and places that both are likely to get you what you want and actually will get you what you want. But there would be no reason to prefer the third list to the first list, given what you care about.
A property P can therefore be good in itself but become irrelevant when conjoined with, say, truth. In this case, we say that P has been swamped by truth. A property whose main claim to fame is to be conducive to truth is hence very likely to be swamped – no one needs to know of something that is true that it is likely to be true. Property P has to be therefore something that has intrinsic or extrinsic value that is not rendered useless if truth is attained. If its value is intrinsic, it is hard to believe it will add anything valuable to true belief if the goal of the game of believing is to attain truth. As Kvanvig (2003: 54) points out, if the added property has the an intrinsic value that makes it good, it cannot compete with truth in its importance for knowledge for “[a]ny claim that there are properties of belief that have value intrinsically, independent of any relationship to the truth, should be met with incredulity.” If its value is extrinsic, it has to be dependent on truth. If it is merely instrumentally dependent, it would also be easily swamped. Therefore, P has to be extrinsically valuable, not merely instrumental to achieve truth and contributing to knowledge if it stands a chance to be both illuminate knowledge and not be swamped by truth.
The issues raised by the value problem hinge directly on the nature of doxastic justification (or of its functional equivalents). What is expected of property P is often believed to be at least partially met by justification. It is interesting to notice quickly that the class of value problems extends to what makes knowledge more invaluable than justified true belief – provided Gettier examples that both classes are not co-extensive. A non-standard response to Gettier examples is to claim that what is really valuable is mere justified true belief. In any case, the issue raised by these value problems is whether anything can in fact make a true belief better. This applied both to adding justification to a true belief and to adding something else to the mix to address Gettier’s examples. Whatever is added needs to satisfy the requisites for P above. Part of the problem of showing how true beliefs can be bettered is to show how justification, or anything similar, adds value to a true belief by being something like P.
This is where things get difficult for externalists. The most well-known form of externalism, which is also a prototypical form of externalism is (process) reliabilism. The necessary condition for knowledge there is to be a reliably acquired true belief. The value problem emerges immediately. As Zagzebski (2003: 13) says a
[…] reliable expresso maker is good because expresso is good. A reliable water-dripping faucet is not good because dripping water is no good. The good of the product makes the reliability of the source that produces it good, but the reliability of the source does not then give the product an additional boost in value. The liquid in this cup is not improved by the fact that it comes from a reliable expresso maker. If the expresso tastes good, it makes no difference if it comes from an unreliable machine. […] If the belief is true, it makes no difference if it comes form an unreliable belief-producing source.
Reliability seems to be conducive to truth but it is not sufficiently non-instrumental not to be swamped by truth. Why would a reliably produced truth be better than its unreliably produced counterpart? Again, as Kvanvig says, if the goal is truth, reliability makes no difference if truth is attained. On the other hand, if it is not, reliability doesn’t make explicit what would the additional goal of knowledge be. In other words, as Swinburne (1999) and Kvanvig (2003) himself agree, reliability has no means to avoid being swamped by truth.
Zagzebski is clear that the value problem for process reliabilism may not generalize to all forms of reliabilism. For instance, agent reliabilism could offer a way to make the reliability of the knower some sort of intellectual virtue that is not merely conducive to truth but invaluable in itself. Moreover, Zagzebski’s own conception of knowledge understands it in terms of intellectual virtue and she claims (in 1996: 299) that intellectual virtue has externalist and internalist elements for one can be intellectually virtuous without knowing it (without accessing her own intellectual virtues). Still, sheer reliabilism – or process reliabilism – seems to be in trouble with the swamping value problem. Why do people, asks Zagzebski in her discussion (1996: 304), think reliable beliefs are both valuable and attach value to knowledge? She thinks that Gettier cases made people believe that knowledge is nonaccidental true belief and reliability makes a belief less accidental. The trouble is, for her, that it provides the wrong type of nonaccidentality. The idea is that if anything external to the believer (like process reliabilism but not agent reliabilism postulates) is called to play the role of property P, it is going to fail not to be swamped by truth. This is because nothing external can seem to add any value to true beliefs that is not merely a capacity to lead towards the truth and if it is so it becomes irrelevant on the face of truth. It is only if there is something at least partially internal that we can hope to protect the additional value composing knowledge from swamping. Kvanvig (2003: 75) makes a similar point explicitly point explicitly for he claims that
[…] we end with a somewhat surprising conclusion: that only a subjective, internalist theory of justification has the capacity to account for the value of justification and avoid the swamping problem.
This sort of justification, even in the presence of true belief, adds value to the composite in question.
It seems like (process) reliabilism is seriously threatened by value problems and further that the treat seems to extend to various other varieties of externalism. Both Kvanvig and Zagzebski believe that anything external to the inner workings of a knower cannot contribute to knowledge by any other means than truth – for any other contribution would be either redundant on the face of truth or utterly irrelevant for the value of knowledge. Zagzebski (2003: 14-16) draws three morals from the value problems:
1. Truth plus a reliable source of truth cannot explain the value of knowledge.
2. Truth plus an independently valuable source cannot explain the value of knowledge.
3. Knowing is related to the knower not as product to machine, but as act to agent.
To be sure, she examines non-externalist accounts of knowledge that have to do with true beliefs where the subject gets credit for the acquired truth (something similar has been proposed by Sosa and Greco). She claims that an expresso machine doesn’t make the coffee better because it gets credit for each coffee it makes. Her point, to do with her third moral above, is that a knower has to have some virtue in order to get credit – the credit itself doesn’t make knowledge more valuable while something that deserves it would. Knowledge could be more valuable than true belief if it is some sort of admirable belief. The challenge to externalism is whether it can provide any sort of account of knowledge that would allow for the knower to become more virtuous or somehow admirable.
((The understanding turn in epistemology, I guess, is also in general a move that favors internalism – it tends to stress the transparency of understanding (as opposed to knowledge). The move towards understanding seems to be a move towards achievement as a value and not as mere success.))
Epistemological disjunctivism
Epistemological disjunctivism is, at the face of it, is a form of externalism. It is a view associated mostly to perceptual knowledge and, as such, it is an alternative to the sort of indirect perception (formulated by Locke and turned more internalist later) that postulates that there is a subjective intermediary between a perceiver and a perceived object, an intermediary that is the same in a case of veridical perception and in a case of hallucination. Disjunctivism is set against ideas around indirect perception that postulates a subjective maximum common factor – to use the term used by McDowell (1982) - is an internal object common to both cases and perceptual justification could make appeal to nothing beyond it. The maximum common content is as far as perceptual justification could go. Disjunctivism posits that there is no such content and perception could go all the way (directly) to the (external) perceived object. In fact, there is no perceived object but the one that is ultimately perceived in veridical cases – no intermediaries, no subjective ideas more directly viewed within the mind. If the object is reached, perceptual justification is appropriate for there is a direct contact between the perceiver and what is perceived. If, however, the object is not there, a completely different story has to be told. In the veridical case, justification arises from the perceiver going all the way towards the perceived object. In a hallucination, there is no justification for there is no object to be perceived. Clearly, the perceiver cannot tell in which case she is while in the exercise of perceiving – justification depends on circumstances unknown to the perceiver. In the terms of Lehrer & Cohen (1983), a new evil genius situation is denied for a (veridical) perceiver and her counterpart whose brain is in a vat are not at the same justificatory status. The veridical perceiver is in a better position but she cannot discriminate from her experience a better and a worse justificatory positions.
Pritchard (2011, 2012) claims that disjunctivism is a holy grail of epistemology for it combines appropriately what is desirable in both the internalist and the externalist camp. He argues that such a combined position is in fact a form of non-classical epistemic internalism (2011: 241). This is because, although the new evil genius case is rejected – for the perceiver cannot discriminate between justified and unjustified cases -, there is an internalist requirement of access that is fulfilled. The requirement is presented as follows (2011: 238):
ACCESS: If S and S* do not differ in the facts that they are able to know by reflection alone then they will not differ in the degree of epistemic justification that they have for their beliefs.
That is, if two perceivers know facts by reflection alone – that is, through perception – they are in the same epistemic state. Understanding access in terms of reflection alone, Pritchard allows for the two perceivers to differ in their own assessment of their epistemic status while being both equally justified, provided that they access facts through perception. Disjunctivism, but not reliabilism, can therefore fulfill the ACCESS requirement that would be enough, for Pritchard, to make it into a variety of (non-classical) internalism. Interestingly, while classical internalism can allow for access through reflection alone to no more than one’s own inner state (mental content, or some form of maximum common factor), disjunctivism allows the perceiver to access external objects through reflection alone because in the veridical case perception goes all the way to the external fact. Access is not confined to an inner circle, it can go all the way to what is outside – access through reflection alone can reach whatever is perceived.
Now, I don’t believe disjunctivism is a genuine form of internalism, classical or non-classical. In the absence of the capacity of the perceiver to discriminate whether or not a genuine justification is available, externalism appears compellingly close. If this is right, Pritchard’s attempt to make disjunctivism a sort of externalism-enhanced variety of internalism fails and we are still navigating in externalist waters. However, he has shown that there could be important differences between externalisms like reliabilism and externalisms like disjunctivism. The difference is established by making a distinction between ACCESS and the capacity to discriminate genuine justifications. Moreover, Pritchard’s claim that disjunctivism is a holy grail can still be somehow justified: disjunctivism could prove to have exactly the appropriate resources to deal with the value problems that have been hard for other forms of externalism. If this is so, it can reconcile the internalist demand for an extrinsically valuable and yet not merely instrumental property to distinguish knowledge and mere true belief with externalist demand for further participation of external elements in the process of attaining knowledge.
Access and perception as achievement
The reconciling trick can be done through the acceptance of ACCESS. Perceptual beliefs are acquired through reflection alone, yet it is an exercise of engaging conceptual abilities to bring in features of the world. As McDowell (1994) puts it, perception is a passive deployment of conceptual reflection devices – perception episodes are episodes of receptivity where the world is in view through the same capacities that are active in thinking. To perceive is therefore an achievement – something akin to an intellectually virtuous act – for one engages one’s reflective devices to capture what is out there. One can fail to capture anything at all, but if one doesn’t, there is a genuine access to the world through an exercise of reflection alone. Perception provides justification because it enables direct involvement with the object and it is valuable because it follows from capacities to detect these objects. Episodes of perceptual knowledge are virtuous because they follow from detecting capacities about which one deserves credit. Surely, perception is neither a route to certainty nor a sure path towards knowledge – as Pritchard´s (2012) enlightening analysis of seeing and knowing makes clear – but it is a valuable justification-contributor. Its value is extrinsic but it cannot be reduced to its instrumentality to tracking truth.
Moreover, perception is a virtue much in the sense Zabzebski (1996, 2003) describes. In fact, if we follow Pritchard’s characterization of disjunctivism as both maintaining the requisite of reflective access and rejecting the requisite of discrimination, disjunctivism, just like Zagzebski’s account of knowledge as intellectual virtue, has elements that enable the knower to be considered virtuous. She points out that her account to combine internalist and externalist elements, as intellectual virtue is something that is not necessarily discriminated by the subject. She reckons her position leaves the debate open as both positions may be adopted (1996: 330). In fact, it seems that one can be unknowingly virtuous as much as one could be unknowingly having an episode of veridical perception – so the new evil genius wouldn’t apply to either case for similar internal states can be in different justification conditions. Perceptual disjunctivism, if we are right about the import of its denial of new evil genius situations, is arguably a more robust form of externalism while maintaining ACCESS and making perception itself invaluable.
H, the analogy with Zagzebski’s position illustrates how promisingly general a disjuntivist path could be. If intellectual value is something the knower cannot discriminate and is crucial for her justificational state, then it seems like the very structure of a Zagzebskian externalism would be disjunctivist. To be sure, it won’t be a perceptual disjunctivism, but still it would be such that something like ACCESS is combined with the knower’s inability to appreciate her justificatory state. This is a general scheme for an externalism that can deal with the value of knowledge problems. Pritchard’s scheme for disjunctivism is, in this sense, a holy grail at least for externalism, for it seems that it can be extended to other varieties of externalism that would make room for a strong tie between reflective access and justification. Including a Zagzebskian variety of externalism.
I conclude with some brief remarks on extending epistemic disjunctivism beyond the purely human cases of perception. These remarks are no more than promissory notes. If perception is an achievement, it is admirable also for non-human episodes of perception that wouldn’t be normally considered intellectually virtuous cases or the output of conceptual capacities. An accurate detection of a salient feature of the environment by, say, a bat is admirable independently of the adequacy of the animal’s representation. Similarly, a hunting animal can perform invaluable perceptive acts even without success in capturing the prey. Domestic animals are often praised for empathy to their human neighbors because they perceive a change in mood through human gestures – even if their wrong about what’s going on with these humans. Pets are empathetic because they are perceptive – and they somehow get credit for this. Finally, there is a sense in which a machine – for instance the one that detects the motive for a baby to cry by being exposed to the sounds the baby does – performs admirable acts of perception that achieve to access salient features of the environment, what is invaluable even if the ultimate perception is not true.
Epistemic externalism tends to make knowledge less dependent on the wonderings of a (human) knowing subject and more as an something attained in collaboration with the environment of the knower. Knowledge is less of a matter of inner cogitations and more an issue of interaction with what is to be known and its surroundings. In order to portray knowledge like that, knowledge-bearers such as beliefs (or opinions, convictions) have to be conceived as less than self-conscious states. However, as truth is normally conceived as being an external contributor to knowledge, the main focus of epistemic externalism has been further clauses that would distinguish knowledge from mere true belief. This means to discuss doxastic justification – or warrant, or truth-conducive procedures – and do it in a way that is not committed to the thesis that the knower has to be able to discriminate genuine cases of justification from deceiving appearances. In other words, in these provinces, the externalist endeavors therefore to show how there could be justification-contributors of which the subject of knowledge is not even potentially aware. One’s experience can be the same as somebody else and she can be more epistemically justified for reasons that are beyond her discrimination capacities.
In askeptical periods – those when, according to Linda Zagzebski’s (2001) terminology, the treat of skepticism is seen as dissolved or somehow innocuous – to investigate knowledge is no mainly focused on proving its possibility but rather on relating it to other notions and placing it among other phenomena. If the skeptic takes for granted that knowledge (or justification) is invaluable and wonders how it can be attained, in askeptical periods we are more likely to wonder why after all knowledge is to be sought as a goal, no matter if it is attainable. Epistemic externalism understands knowledge as firmly placed in the world around the knower rather than taking the knower to be to a large measure acting (or judging) from afar. Externalism is also akin to the idea that knowledge is not merely human, or at least not in principle dependent on anything specifically human. If knowledge has a value, it has to be somehow a value that is not tied to specific human faculties nor associated to distinctively human purposes. The value of knowledge would have to be construed not only out of something only humans can appreciate.
The value problem for externalists
In fact, the value problem has become a major problem for externalists recently. The general problem – or rather the class of problems – is old and familiar. In Meno, Plato questions whether knowledge is itself more invaluable than mere true belief. After all, in most cases, if one’s beliefs are true – about how to cook a mole or how to go back to one’s hotel in a Southern Hemisphere town – it is for most purposes enough. It is clear that it is not enough to come up with a property P that is itself invaluable and distinguishes knowledge from true belief for P can add no value to the mix. One has rather to consider how P would interact with a true belief. Kvanvig (2003: 45) explains why:
[…] consider some simple analogies. If we have a piece of art that is beautiful, its aesthetic value is not enhanced by having as well the property of being likely to be beautiful. For being likely to be beautiful is a valuable property because of its relationship to being beautiful itself. Once beauty is assumed to be present, the property of being likely to be beautiful ceases to contribute […] Take anything that you care about: happiness, money, drugs, sports cars, and so on. Then consider two lists about such things, the first list telling you where to obtain such things and the second list telling you where you are likely to obtain such things. Now compose a third list, which is the intersection of the first two lists. It tells you of ways and places that both are likely to get you what you want and actually will get you what you want. But there would be no reason to prefer the third list to the first list, given what you care about.
A property P can therefore be good in itself but become irrelevant when conjoined with, say, truth. In this case, we say that P has been swamped by truth. A property whose main claim to fame is to be conducive to truth is hence very likely to be swamped – no one needs to know of something that is true that it is likely to be true. Property P has to be therefore something that has intrinsic or extrinsic value that is not rendered useless if truth is attained. If its value is intrinsic, it is hard to believe it will add anything valuable to true belief if the goal of the game of believing is to attain truth. As Kvanvig (2003: 54) points out, if the added property has the an intrinsic value that makes it good, it cannot compete with truth in its importance for knowledge for “[a]ny claim that there are properties of belief that have value intrinsically, independent of any relationship to the truth, should be met with incredulity.” If its value is extrinsic, it has to be dependent on truth. If it is merely instrumentally dependent, it would also be easily swamped. Therefore, P has to be extrinsically valuable, not merely instrumental to achieve truth and contributing to knowledge if it stands a chance to be both illuminate knowledge and not be swamped by truth.
The issues raised by the value problem hinge directly on the nature of doxastic justification (or of its functional equivalents). What is expected of property P is often believed to be at least partially met by justification. It is interesting to notice quickly that the class of value problems extends to what makes knowledge more invaluable than justified true belief – provided Gettier examples that both classes are not co-extensive. A non-standard response to Gettier examples is to claim that what is really valuable is mere justified true belief. In any case, the issue raised by these value problems is whether anything can in fact make a true belief better. This applied both to adding justification to a true belief and to adding something else to the mix to address Gettier’s examples. Whatever is added needs to satisfy the requisites for P above. Part of the problem of showing how true beliefs can be bettered is to show how justification, or anything similar, adds value to a true belief by being something like P.
This is where things get difficult for externalists. The most well-known form of externalism, which is also a prototypical form of externalism is (process) reliabilism. The necessary condition for knowledge there is to be a reliably acquired true belief. The value problem emerges immediately. As Zagzebski (2003: 13) says a
[…] reliable expresso maker is good because expresso is good. A reliable water-dripping faucet is not good because dripping water is no good. The good of the product makes the reliability of the source that produces it good, but the reliability of the source does not then give the product an additional boost in value. The liquid in this cup is not improved by the fact that it comes from a reliable expresso maker. If the expresso tastes good, it makes no difference if it comes from an unreliable machine. […] If the belief is true, it makes no difference if it comes form an unreliable belief-producing source.
Reliability seems to be conducive to truth but it is not sufficiently non-instrumental not to be swamped by truth. Why would a reliably produced truth be better than its unreliably produced counterpart? Again, as Kvanvig says, if the goal is truth, reliability makes no difference if truth is attained. On the other hand, if it is not, reliability doesn’t make explicit what would the additional goal of knowledge be. In other words, as Swinburne (1999) and Kvanvig (2003) himself agree, reliability has no means to avoid being swamped by truth.
Zagzebski is clear that the value problem for process reliabilism may not generalize to all forms of reliabilism. For instance, agent reliabilism could offer a way to make the reliability of the knower some sort of intellectual virtue that is not merely conducive to truth but invaluable in itself. Moreover, Zagzebski’s own conception of knowledge understands it in terms of intellectual virtue and she claims (in 1996: 299) that intellectual virtue has externalist and internalist elements for one can be intellectually virtuous without knowing it (without accessing her own intellectual virtues). Still, sheer reliabilism – or process reliabilism – seems to be in trouble with the swamping value problem. Why do people, asks Zagzebski in her discussion (1996: 304), think reliable beliefs are both valuable and attach value to knowledge? She thinks that Gettier cases made people believe that knowledge is nonaccidental true belief and reliability makes a belief less accidental. The trouble is, for her, that it provides the wrong type of nonaccidentality. The idea is that if anything external to the believer (like process reliabilism but not agent reliabilism postulates) is called to play the role of property P, it is going to fail not to be swamped by truth. This is because nothing external can seem to add any value to true beliefs that is not merely a capacity to lead towards the truth and if it is so it becomes irrelevant on the face of truth. It is only if there is something at least partially internal that we can hope to protect the additional value composing knowledge from swamping. Kvanvig (2003: 75) makes a similar point explicitly point explicitly for he claims that
[…] we end with a somewhat surprising conclusion: that only a subjective, internalist theory of justification has the capacity to account for the value of justification and avoid the swamping problem.
This sort of justification, even in the presence of true belief, adds value to the composite in question.
It seems like (process) reliabilism is seriously threatened by value problems and further that the treat seems to extend to various other varieties of externalism. Both Kvanvig and Zagzebski believe that anything external to the inner workings of a knower cannot contribute to knowledge by any other means than truth – for any other contribution would be either redundant on the face of truth or utterly irrelevant for the value of knowledge. Zagzebski (2003: 14-16) draws three morals from the value problems:
1. Truth plus a reliable source of truth cannot explain the value of knowledge.
2. Truth plus an independently valuable source cannot explain the value of knowledge.
3. Knowing is related to the knower not as product to machine, but as act to agent.
To be sure, she examines non-externalist accounts of knowledge that have to do with true beliefs where the subject gets credit for the acquired truth (something similar has been proposed by Sosa and Greco). She claims that an expresso machine doesn’t make the coffee better because it gets credit for each coffee it makes. Her point, to do with her third moral above, is that a knower has to have some virtue in order to get credit – the credit itself doesn’t make knowledge more valuable while something that deserves it would. Knowledge could be more valuable than true belief if it is some sort of admirable belief. The challenge to externalism is whether it can provide any sort of account of knowledge that would allow for the knower to become more virtuous or somehow admirable.
((The understanding turn in epistemology, I guess, is also in general a move that favors internalism – it tends to stress the transparency of understanding (as opposed to knowledge). The move towards understanding seems to be a move towards achievement as a value and not as mere success.))
Epistemological disjunctivism
Epistemological disjunctivism is, at the face of it, is a form of externalism. It is a view associated mostly to perceptual knowledge and, as such, it is an alternative to the sort of indirect perception (formulated by Locke and turned more internalist later) that postulates that there is a subjective intermediary between a perceiver and a perceived object, an intermediary that is the same in a case of veridical perception and in a case of hallucination. Disjunctivism is set against ideas around indirect perception that postulates a subjective maximum common factor – to use the term used by McDowell (1982) - is an internal object common to both cases and perceptual justification could make appeal to nothing beyond it. The maximum common content is as far as perceptual justification could go. Disjunctivism posits that there is no such content and perception could go all the way (directly) to the (external) perceived object. In fact, there is no perceived object but the one that is ultimately perceived in veridical cases – no intermediaries, no subjective ideas more directly viewed within the mind. If the object is reached, perceptual justification is appropriate for there is a direct contact between the perceiver and what is perceived. If, however, the object is not there, a completely different story has to be told. In the veridical case, justification arises from the perceiver going all the way towards the perceived object. In a hallucination, there is no justification for there is no object to be perceived. Clearly, the perceiver cannot tell in which case she is while in the exercise of perceiving – justification depends on circumstances unknown to the perceiver. In the terms of Lehrer & Cohen (1983), a new evil genius situation is denied for a (veridical) perceiver and her counterpart whose brain is in a vat are not at the same justificatory status. The veridical perceiver is in a better position but she cannot discriminate from her experience a better and a worse justificatory positions.
Pritchard (2011, 2012) claims that disjunctivism is a holy grail of epistemology for it combines appropriately what is desirable in both the internalist and the externalist camp. He argues that such a combined position is in fact a form of non-classical epistemic internalism (2011: 241). This is because, although the new evil genius case is rejected – for the perceiver cannot discriminate between justified and unjustified cases -, there is an internalist requirement of access that is fulfilled. The requirement is presented as follows (2011: 238):
ACCESS: If S and S* do not differ in the facts that they are able to know by reflection alone then they will not differ in the degree of epistemic justification that they have for their beliefs.
That is, if two perceivers know facts by reflection alone – that is, through perception – they are in the same epistemic state. Understanding access in terms of reflection alone, Pritchard allows for the two perceivers to differ in their own assessment of their epistemic status while being both equally justified, provided that they access facts through perception. Disjunctivism, but not reliabilism, can therefore fulfill the ACCESS requirement that would be enough, for Pritchard, to make it into a variety of (non-classical) internalism. Interestingly, while classical internalism can allow for access through reflection alone to no more than one’s own inner state (mental content, or some form of maximum common factor), disjunctivism allows the perceiver to access external objects through reflection alone because in the veridical case perception goes all the way to the external fact. Access is not confined to an inner circle, it can go all the way to what is outside – access through reflection alone can reach whatever is perceived.
Now, I don’t believe disjunctivism is a genuine form of internalism, classical or non-classical. In the absence of the capacity of the perceiver to discriminate whether or not a genuine justification is available, externalism appears compellingly close. If this is right, Pritchard’s attempt to make disjunctivism a sort of externalism-enhanced variety of internalism fails and we are still navigating in externalist waters. However, he has shown that there could be important differences between externalisms like reliabilism and externalisms like disjunctivism. The difference is established by making a distinction between ACCESS and the capacity to discriminate genuine justifications. Moreover, Pritchard’s claim that disjunctivism is a holy grail can still be somehow justified: disjunctivism could prove to have exactly the appropriate resources to deal with the value problems that have been hard for other forms of externalism. If this is so, it can reconcile the internalist demand for an extrinsically valuable and yet not merely instrumental property to distinguish knowledge and mere true belief with externalist demand for further participation of external elements in the process of attaining knowledge.
Access and perception as achievement
The reconciling trick can be done through the acceptance of ACCESS. Perceptual beliefs are acquired through reflection alone, yet it is an exercise of engaging conceptual abilities to bring in features of the world. As McDowell (1994) puts it, perception is a passive deployment of conceptual reflection devices – perception episodes are episodes of receptivity where the world is in view through the same capacities that are active in thinking. To perceive is therefore an achievement – something akin to an intellectually virtuous act – for one engages one’s reflective devices to capture what is out there. One can fail to capture anything at all, but if one doesn’t, there is a genuine access to the world through an exercise of reflection alone. Perception provides justification because it enables direct involvement with the object and it is valuable because it follows from capacities to detect these objects. Episodes of perceptual knowledge are virtuous because they follow from detecting capacities about which one deserves credit. Surely, perception is neither a route to certainty nor a sure path towards knowledge – as Pritchard´s (2012) enlightening analysis of seeing and knowing makes clear – but it is a valuable justification-contributor. Its value is extrinsic but it cannot be reduced to its instrumentality to tracking truth.
Moreover, perception is a virtue much in the sense Zabzebski (1996, 2003) describes. In fact, if we follow Pritchard’s characterization of disjunctivism as both maintaining the requisite of reflective access and rejecting the requisite of discrimination, disjunctivism, just like Zagzebski’s account of knowledge as intellectual virtue, has elements that enable the knower to be considered virtuous. She points out that her account to combine internalist and externalist elements, as intellectual virtue is something that is not necessarily discriminated by the subject. She reckons her position leaves the debate open as both positions may be adopted (1996: 330). In fact, it seems that one can be unknowingly virtuous as much as one could be unknowingly having an episode of veridical perception – so the new evil genius wouldn’t apply to either case for similar internal states can be in different justification conditions. Perceptual disjunctivism, if we are right about the import of its denial of new evil genius situations, is arguably a more robust form of externalism while maintaining ACCESS and making perception itself invaluable.
H, the analogy with Zagzebski’s position illustrates how promisingly general a disjuntivist path could be. If intellectual value is something the knower cannot discriminate and is crucial for her justificational state, then it seems like the very structure of a Zagzebskian externalism would be disjunctivist. To be sure, it won’t be a perceptual disjunctivism, but still it would be such that something like ACCESS is combined with the knower’s inability to appreciate her justificatory state. This is a general scheme for an externalism that can deal with the value of knowledge problems. Pritchard’s scheme for disjunctivism is, in this sense, a holy grail at least for externalism, for it seems that it can be extended to other varieties of externalism that would make room for a strong tie between reflective access and justification. Including a Zagzebskian variety of externalism.
I conclude with some brief remarks on extending epistemic disjunctivism beyond the purely human cases of perception. These remarks are no more than promissory notes. If perception is an achievement, it is admirable also for non-human episodes of perception that wouldn’t be normally considered intellectually virtuous cases or the output of conceptual capacities. An accurate detection of a salient feature of the environment by, say, a bat is admirable independently of the adequacy of the animal’s representation. Similarly, a hunting animal can perform invaluable perceptive acts even without success in capturing the prey. Domestic animals are often praised for empathy to their human neighbors because they perceive a change in mood through human gestures – even if their wrong about what’s going on with these humans. Pets are empathetic because they are perceptive – and they somehow get credit for this. Finally, there is a sense in which a machine – for instance the one that detects the motive for a baby to cry by being exposed to the sounds the baby does – performs admirable acts of perception that achieve to access salient features of the environment, what is invaluable even if the ultimate perception is not true.
Comments
Post a Comment