Skip to content

Fr. 467

December 24, 2013

The limiting case now in place, how do Goldman and Olsson go about accounting for knowledge’s surplus value? The surplus value problem proper to the reliabilist framework is that of “the swamping argument” (to which Greco makes passing reference without explicitly naming it). Although this is not the place to enter into the details of such an account, we will give some general indications of the argument’s basic claims. Insofar as knowledge’s value, as per reliabilism, stems from the truth-conduciveness of epistemic and justificatory processes, it might then seem that these processes’ value also owes to truth. Furthermore, if the value of merely true beliefs also owes to truth, then the value of both knowledge and merely true beliefs stems from their truth. As that which determines the value of knowledge is precisely that which determines the value of merely true beliefs, it would follow that knowledge, despite its additional components, brings no new value to the target proposition than that already offered by merely true belief. In short, as per a common formulation of the swamping argument, knowledge’s value is parasitic on that of merely true belief. (For a more detailed account of the swamping argument, see “Problems for Early Process Reliabilism” in the entry for “Reliabilism” in the Stanford Encyclopedia of Philosophy, as well as Goldman and Olsson, pp. 23-27.)

Rather than a closer examination of Goldman and Olsson’s breakdown and diagnosis of this objection, we shall focus on the solutions from a reliabilist framework proposed by the authors to this objection. These solutions are the following: the conditional probability solution and the value autonomization solution. As noted in the text, these solutions are not mutually exclusive and may even prove complementary, one entailing the other.

Of the first, the authors begin with the supposition that knowledge on a reliabilist account produces a state of affairs including a certain subset of elements, e.g. justification through a reliable process, true belief, an anti-Gettier condition, and so on. Among these elements is to be found a property that is not present in cases of non-knowledge (or merely true belief) on the strong reading, be it of the 1.), 2.) or 3.) variety. For this reason, this property must be epistemically valuable, sets knowledge apart from merely true belief, and is, therefore, that which makes knowledge more valuable than merely true belief. On a reliabilist account, this property consists in “making it likely that one’s future beliefs of a similar kind will also be true” (Goldman and Olsson, p. 28, emphasis in the original). Put somewhat differently, the likelihood of arriving at similar true beliefs in the future is greater in virtue of arriving at a generalizable true belief through knowledge and the reliable epistemic processes entailed than through merely true belief and the illegitimate justification or blind luck that this implies. Hence, the name of the solution: the probability that a belief is generalizable to other cases hinges on or is conditional upon its having been arrived at through the proper epistemic means. This way of improving generalizability of a target belief is, without a doubt, valuable. (The term generalizability is ours and, at a glance, emcompasses the empiricial regularities that Goldman and Olsson introduce below.)

Yet this solution remains silent on the precise characterization of this value, a fact that Goldman and Olsson count to its credit. For this lack of outright commitment permits the conditional probability solution to sidestep the swamping argument by singling out the presence of a valuable property, whatever its value, in the composite state of affairs that attaches to knowing (in the strong sense) and which is otherwise absent from cases of merely true belief. In sum, the value of this property is absent from the merely true belief state, being present only in the knowledge state, and thus demonstrates that knowledge does have some value above and beyond that of merely true belief.

This argumentative move does, however, leave one with the sensation that the account of knowledge’s value is incomplete so long as the kind of value at issue remains unspecified. Indeed, this seems precisely the sort of demand that the instrumental/intrinsic dilemma places on the epistemologist. To return to the present account, although we might tend to characterize the goal of improving generalizability and the values that accrues to it as broadly instrumental in scope, there is no explicit move on the part of the authors to this conclusion. While the tendency might be to think of such improvement in terms of identifying the proper means to solve current and future problems, thus suggesting instrumental value, there is no prima facie reason to think that wisdom with its intrinsic value does not also stand to benefit from generalizability or broader application to a variety of scenarios. It also bears mentioning that postulating an indeterminate surplus value for knowledge does not necessitate that this value is by its nature indeterminate, but simply that it has yet to be specified. For the moment, it is enough to show that there is some plausible account of surplus value.

The groundwork in place, Goldman and Olsson return to the notion of limiting cases above in order to show that, although this reliabilist account of knowledge’s value groups a large number of cases, there is no need for it to account for each and every instance of identifiable knowledge. Specifically, this solution is subject to restrictions in that generalizability of the sort above hinges on certain “empirical regularities” (ibid., p. 29). These are four in number:

1.) non-uniqueness
2.) cross-temporal access
3.) learning
4.) generality

It is only in cases where these four conditions are united that there can be surplus value attributed to knowledge, i.e. values that stems specifically from the conditional probability account provided above. In other words, knowledge of a given target proposition is valuable to the degree “to which the assumptions of non-uniqueness, cross-temporal access, learning and generality are satisfied in a given case” (ibid., p. 30). In cases where these regularities are not present, generalizability cannot be the aim of the justificatory processes as the most salient features of the situation or target proposition do not lend themselves to this end. From the above, it follows, of course, that, in the absence of these regularities, knowledge cannot be said to have surplus value in relation to merely true belief. In short, the present solution does not postulate surplus value in all cases.

Does this leave the authors’ first solution fatally flawed? This is only the case insofar as the opponent of surplus value maintains that surplus value accounts of knowledge must show that knowledge always has this extra value. Whence the importance of proving the existence of a weak sense of knowledge for Goldman and Olsson. So long as there is a plausible and widely accepted case in which knowledge’s value is equal to that of merely true belief and exceptions to surplus value do not violate the viability of a limited account of knowledge’s value, there is no reason to suppose that a solution to the surplus value problem need involve an unqualified claim that knowledge is more valuable in each and every case. Although the epistemological community is divided on this necessity of such a claim, Goldman and Olsson find sufficient reason for a qualified account in the fact that generalizations are by their nature an affair of tendencies rather than the absolute. After all, “the generalizations we make in our daily lives are not universal generalizations in the sense of predicate logic but elastic generic claims that can survive a limited number of counter-instances”, thus suggesting that, for the purposes of empirical or practical generality, there is no need to adhere to the strictures of conceptual generality (ibid., p. 31).

(Interestingly, these brief indications from the authors seem to fall under the umbrella of the second negative argumentative strategy highlighted in the first part of this examination in that they here seek to dispel confusions surrounding the generality proper to the surplus value problem. For this reason, the authors might be considered to endorse an overall positive argumentative strategy that nonetheless makes use of both negative elements to motivate its claims.)

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: