Anchors on Neurath’s Boat: Non-Foundationalist Epistemic Entitlements Jonathan Lopez

advertisement
Volume 1 (2013)
Anchors on Neurath’s Boat: Non-Foundationalist Epistemic Entitlements
Jonathan Lopez
Department of Philosophy, University of British Columbia
Abstract
Recent developments in epistemology have attempted to revive the foundationalist picture of knowledge
with the notion of epistemic entitlement. Essentially, epistemic entitlement is the idea that we are entitled
to some beliefs, that is, we do not have to earn warrant in order to believe them, they come to us for free.
Wright explicitly attempts to argue that beliefs we are entitled to, cornerstone propositions, can serve as
foundations for our knowledge. A related account of epistemic entitlement, although not with the explicit
aim of finding foundations, is also considered. I argue that Wright is successful on pragmatic grounds
rather than the more salient and desired epistemic grounds. Falling short of his goals, Wright’s notion of
epistemic entitlement can be recast into a non-foundationalist picture. The prospect of applying Wright’s
idea of epistemic entitlement to Quine’s web of belief is explored.
I. Introduction
A tempting way of addressing the Cartesian skeptic, some might say intuitive, is to offer some basic
belief to stop the skeptic’s regress argument (i.e. that the justification for one’s justification could be
demanded ad infinitum). A crucial aim of the epistemic entitlement movement is an attempt to find
foundational regress stoppers which require no further justification. The aim of this paper will be to
examine and evaluate the extent to which entitlements can be thought to serve as foundations for our
knowledge. Further, I will argue that notions of epistemic entitlement, falling short of providing
foundations for knowledge, provide us with important insight about how we can anchor knowledge rather
than placing it on foundations. The difference being that in anchoring our knowledge we only make it
more secure than it previously was, perhaps helpful during revisions, rather than placing it on firm ground
as the foundationalist would like. To this end I will sketch out how notions of epistemic entitlement might
be recast into the non-foundationalist picture.
The first two sections of this paper are dedicated to evaluating two notions of epistemic
entitlement that either explicitly or implicitly adopt a foundationalist framework to stave off skeptical
44
Jonathan Lopez/Volume 1 (2013)
worries. Attention will then turn to exploring how epistemic entitlements could be recast into a nonfoundational picture. The primary focus will be on Wright’s notion of epistemic entitlement because of
striking similarities with Quine’s idea of a web of belief. The third section of this paper will be dedicated
to detailing these similarities. The final sections of the paper are dedicated at examining what role
epistemic entitlements can play in a non-foundational framework.
II. A “Too Big to Fail” Interpretation of Wright’s Epistemic Entitlement
One way of understanding Wright’s epistemic entitlement project is an attempt to block skeptical worries,
such as “how do I know there is an external world” or “how am I justified in using induction”, by
providing firm grounds or foundations upon which to rest knowledge. The firm grounds or foundations
being sought here is in the foundationalist spirit that wants to base knowledge on indubitable basic
beliefs. To this end Wright proposes a ‘cornerstone’ for a given region of thought which he takes to be
something such that lacking it would mean you could not rationally claim warrant for any belief in that
region.1 Put another way, if someone has built up a large corpus of knowledge around a proposition like “trust one’s senses” - then one is entitled to this ‘cornerstone’ proposition as otherwise they would lose
warrant for any claim in that area.
Wright then begins to describe two of the kinds of skepticism that have plagued Epistemology.
Cartesian skepticism is concerned with one’s inability to gain warrant for believing we are not being
tricked by an evil demon or that we are not brains in vats. The Humean kind of skepticism works
differently in that it specifically targets one’s ability to make inferences. For example, from the
observation that the sun rises in the morning one might be tempted to say that the sun will rise every
morning and as such the sun will rise tomorrow morning. Hume warns that in order to make this inference
one tacitly assumes the uniformity of nature which roughly maintains that nature will continue to behave
in an orderly fashion. Upon realizing one tacitly invokes the principles of uniformity Hume points out that
one lacks warrant in maintaining this crucial premise.
1
Crispin Wright, “Warrant for Nothing (And Foundations for Free?),” Aristotelian Society Supplementary Volume
78 (2004): 167-212.
45
Jonathan Lopez/Volume 1 (2013)
Generalizing the skeptical problem Wright introduces the I-II-III argument2 to capture the
skeptical worry:
I: The sun has been observed to rise every morning
II: The sun will rise every morning
III: The Uniformity of Nature
Wright points out that even though (II) entails (III) warrant does not transmit. We do, however,
have warrant for cornerstone propositions of type (3); these are unearned warrants or entitlements. By
framing the skeptical arguments as I-II-III arguments Wright points out that their aim is to show that they
target our ability to earn a warrant for type 3 proposition. Wright then turns to explaining why we have
these unearned warrants.
The first of the three kinds of Entitlements Wright introduces (which are examined here3) is
Strategic Entitlement (SE). Wright has us suppose Crusoe, marooned on a deserted island, is starving to
death when he comes across a fruit bearing tree. Crusoe then begins to contemplate whether or not he
should eat the fruit remembering that the fruit could be poisonous. Since Crusoe does not expect to be
rescued any time soon and there is other suitable form of nourishment nearby he reasons if he does not eat
the fruit he will starve to death. However, if he chooses to eat the fruit he will stave off hunger but run the
risk of eating a poisonous fruit and subsequently die. From these considerations Crusoe reasons that he
should eat the fruit because if he doesn't eat the fruit he will surely die whereas he might live if he does
eat the fruit. The foregoing game theoretic analysis would call eating the fruit a dominant strategy
because one sometimes does better but never worse.
The game theoretic example that Wright provided is supposed to be no different than
Reichenbach`s pragmatic vindication of induction. Following a very similar structure, Reichenbach sets
up a decision matrix where it is a dominant strategy to reason according to induction.
2
This argument can be formulated to fit the Cartesian or Humean Skeptic, though only the Humean formulation is
provided.
3
Entitlement of Substance is not explored here because by Wright’s own admission this view is only canvassed.
46
Jonathan Lopez/Volume 1 (2013)
Uniformity of Nature
Nature is not Uniform
Induction
success
failure
Non-Inductive method
success or failure
failure
Success, for Reichenbach, is the ability to make predictions about the future. A non-inductive
method is simply any method that does not rely on induction, such as, coin flipping, going with a gut
feeling, astrology, etc. Any of these methods may, at times, be successful at predicting the future but this
is just as long as nature is uniform. Should nature not behave uniformly then all bets are off for with
respect using a systematic method to predict the future. Put crudely, if nature isn’t going to cooperate, no
method is going to reliably inform you about the future. The important thing to notice here is, similar to
the Crusoe example, there is a dominant strategy which is to use induction.
An important feature to note about the argument given above is that while it might provide an
answer to the Humean skeptic it has nothing to offer the Cartesian skeptic. The Cartesian skeptic would
be dissatisfied because no similar argument can be constructed to accept that there is an external world.
The Humean skeptic is ultimately left disappointed because they already knew that it would be prudent to
accept that nature behaves uniformly, what they’re after is non-pragmatic justification. Put another way,
this is not an epistemic reason for accepting induction, rather it is pragmatic. Wright remarks on the
limitations of the entitlements discussed so far that they presuppose the material world and therefore
won’t “cure ontological skepticism”, doubt of an external world, and are instead aimed at methodological
doubt, reliance on sense perceptions. The other types of epistemic entitlement are meant to address this
worry.
A similar kind of entitlement, Entitlement of Rational Deliberation (ERD) says we are entitled to
place trust in any general presupposition that is conducive to rational decision-making.4 The motivation
behind ERD is that Wright sees rational agency as something we cannot “opt out of”. As such we are
“entitled to place trust [in] whatever we have no evidence against and which needs to be true for rational
4
Crispin Wright, “Warrant for Nothing (And Foundations for Free?),” 197
47
Jonathan Lopez/Volume 1 (2013)
decision making is to be feasible and effective.”5 Basically, the need to take action, will trump any
consideration of skeptical possibility and Wright says inductive and deductive thinking is “the best we
can knowledgeably do in pursuit of purposes which are essential to rational agency.”6 An agent might
employ rational decision-making implicitly in mundane everyday activities such as buying textbooks but
according to ERD they are entitled to them. Say for example, an agent has to buy textbooks and only has
$125 with which to buy two textbooks that cost $68 and $57. While a skeptic might point out that the
basic rules of arithmetic may no longer apply when numbers get this big, the agent reasons that they have
enough money and makes the rational decision to buy the textbooks. ERD would say the agent is entitled
to believing that arithmetic rules still apply and ignore the skeptical possibility because arithmetic is
indispensable in making these kinds of decisions.
There is a familiar pattern with respect to how SE and ERD attempt to respond to skeptical
worries. If successful we would have an entitlement to induction which would answer the Humean but not
the Cartesian skeptic. This is because induction could be thought of as conducive to rational-decision
making (i.e. scientific generalizations) entitling an agent to place trust in such a method. The skeptic
might reply, similarly as he did in the case of SE, that these are pragmatic grounds for the acceptance of a
general proposition on had no warrant for in the first place. If induction is something an agent needs for
rational decision, so much the worse for rational decision making. Wright’s pragmatic motivation shines
brightest when he remarks that “reliance on inductive and abductive methods is the best we can
knowledgeably do in pursuit of purposes which are essential to rational agency” which would hardly
impress the hardline skeptic.7
The most promising kind of entitlement that Wright proposes is Entitlement of Cognitive Project
(ECP) which departs from the pragmatic spirit of the previous two types of entitlements. The conditions
for this kind of entitlement is that the proposition in question be a presupposition of a cognitive project. A
cognitive project, for example, could be using one’s vision to safely navigate highway traffic in order to
accomplish some immensely important task like getting to a hospital. The presupposition here would be
that our vision puts us in proper association with the material objects we perceive. Doubting one’s vision
in this example would lead to a severe “doubting in the competence of the project” but failing to get to the
5
Crispin Wright, “Warrant for Nothing (And Foundations for Free?),” 198
Crispin Wright, “Warrant for Nothing (And Foundations for Free?),” 200
7
Crispin Wright, “Warrant for Nothing (And Foundations for Free?),” 200
6
48
Jonathan Lopez/Volume 1 (2013)
hospital might be a matter of life or death. So one might as well take a chance and drive through highway
traffic, even if our eyesight might be faulty, as not making it to the hospital could be critical. Fortunately,
our vision generally works in a way that when we reach toward objects we are met with resistance rather
than our hand going through objects we perceive to be material which can be taken as evidence that our
presupposition is not untrue. The final condition, is that further presuppositions offer no further epistemic
standing than the original presupposition. This is to ensure that we attribute entitlement to the appropriate
presupposition. Crudely put, if our presupposition requires another presupposition then this will earn
entitlement under ECP. If one can rationally trust their senses to give them true information of the
external world than we can disregard the Cartesian skeptic who would have us believe we might be
dreaming and we are not properly acquainted with the external world. We can simply say to the Cartesian
skeptic that we are entitled via ECP to the presupposition that we are acquainted with the external world
to stop his regress argument. If we can justify our presuppositions via ECP then all we need is a cognitive
project which presupposes the uniformity of nature and we will similarly satisfy the Humean skeptic.
As before it would be important to ensure our entitlements are not reliant on pragmatic grounds.
With respect to ES and ERD it has been shown that Wright invoked pragmatic rather than epistemic
reasons to trust our entitlements. ECP falls short of the mark of counting as a foundation for knowledge
because Wright points out that it would be impossible to check all the presupposition of a cognitive
project which do not put us in a better epistemic position.8 An analogy with ethics can be seen as Wright
is essentially invoking the “ought-implies-can” principle for ECP.9 Roughly speaking, we are not morally
obligated to do something that we are unable to do like achieve world peace tomorrow. We would not be
held responsible if we fail to create world peace in a day’s time because this is something which is
thought to be impossible for a person to undertake, especially under the time constraints. Similarly Wright
wants to say that we shouldn’t be forced into something we simply are not able to do (i.e. check all our
presuppositions ad infinitum). However, this is not meeting the Cartesian skeptic on his own grounds
because they would reject that the ought implies can principle would hold in epistemology. 10 If checking
all our presuppositions is something that humans are unable to do the Skeptic would simply retort “all the
worse for knowledge.” While it may be pragmatic not to check every presupposition (i.e. regress) the
skeptic will still want an epistemic reason to stop the regress argument. To say that it would be too
8
Wright, 189
Carrie Jenkins, “Entitlement and Rationality,” Synthese 157 (2008): 25–45.
10
Carrie Jenkins, “Entitlement and Rationality,” 113–127
9
49
Jonathan Lopez/Volume 1 (2013)
difficult to check all our propositions so we shouldn’t be asked to try would amount to admitting the
skeptic’s point. A possible move for the foundationalist might be to search out a different notion of
epistemic entitlement for foundations.
III. Field Entitlement on Default Reasonableness
In the spirit of Epistemic Entitlement, Field can be interpreted as trying to provide a regress stopper by
appealing to “default reasonableness”. Although Field’s larger project is to give an evaluative account of
a priori knowledge, the purposes of this paper will largely only consider his arguments for why we accept
“default reasonableness”.
Field defines a proposition as default reasonable if “it is reasonable to believe or employ it
without first adducing evidence or arguments in its favour.”11Although Field doesn’t focus on how this
kind of reasonable can fit into a foundationalist picture, it can immediately be appreciated that if we
accept “default reasonable” propositions, then we have something to offer the skeptic to stop their regress
argument. A natural questions to ask at this point is what makes a rule of proposition reasonable and why
would it be our default position.
Field maintains that in calling a rule or a proposition reasonable, all we are doing is evaluating it.
This is very different approach than understanding reasonableness as a factual property which is made
true or false depending on how it stands to a mind-independent feature of the world. So when it comes to
deciding whether something is reasonable or not, all we're doing is reporting our attitude toward some
proposition, revealing only what we value. For the evaluativists, like Field, reliability of our inductive and
perceptual rules are simply a characteristic that we place high value on conducive to us giving it a
`thumbs up`. Field intends to contrast his position with ‘factualists’ who would render a proposition
reasonable or not just in case it meets certain reliability criterion, whatever they may be.
Among the prospects Field sees for reasonableness being understood as an evaluative property is
that it removes the force of most skeptical arguments.12 On Field’s view the hardline skeptic would be
doing no more than asserting his attitude one what should count as reasonable when they say that
11
Hartry Field, "Apriority as an Evaluative Notion," in New Essays on the A Priori, eds. Paul Boghossian and
Christopher Peacocke (New York: Oxford University Press, 2001), 124.
12
Hartry Field, "Apriority as an Evaluative Notion," 143
50
Jonathan Lopez/Volume 1 (2013)
knowledge is impossible. By switching to an Evaluativist picture the judgment that there are no
reasonable beliefs simply become a report of the skeptic’s attitude and the problem is skirted since the
skeptic is not thought to be making a factual statement about the status of knowledge.
There are two reasons that we can reject Field’s ‘default reasonableness’ as serving for
foundations. The first is that by his own admission, Evaluativism is going to allow for at least a moderate
kind of relativism over what one takes to be epistemically valuable.13 For example, if one day you
discover your friend does not enjoy the same ice cream flavor, say cookie dough ice cream, you are not
likely to convince him your flavor is really the best by pointing out how it tastes (i.e. one might cite the
taste of raw cookie dough or chocolate chips). Similarly, if someone simply doesn’t put much value into
reliability of a proposition (perhaps they prefer truth preservation) it doesn’t seem like very persuasive
arguments can be made in either direction under the Evaluativist notion. If someone does not think
reliability is good quality nothing more can be said to them than can be said to someone that does not
share in your favorite ice cream flavor. The skeptic might even go further as to say there is no principled
reason they should even care about achieving truths and avoiding falsehoods thereby placing little to no
value in common features such as reliability or truth preservation. One might try to guard against this by
saying one ought to take some things as valuable but this would require a whole suite of separate
considerations.
Another reason default reasonableness is not going to serve as a foundation for knowledge can be
seen when Field explains “that if one is going to have very many beliefs at all one must have default
rules.”14 He goes on to say to get to this conclusion that some rules are default reasonable, one needs to
assume that it is possible to have a sufficient array of reasonable beliefs; and to get to the conclusion that
some of the rules we employ are default reasonable, one needs to assume that some of our own beliefs are
reasonable. Of course the skeptic would be very unmoved at this point, because if skeptic agreed with the
premise that some beliefs are reasonable then they wouldn't count as skeptics. Put less facetiously, for any
belief that one might be tempted to call reasonable, perhaps “I have a hand”, the skeptic would call into
question and deny its reasonableness. To the determined skeptic it does not seem like this is going to be a
very solid foundation.
13
14
Hartry Field, "Apriority as an Evaluative Notion," 140- 141.
Hartry Field, "Apriority as an Evaluative Notion," 124.
51
Jonathan Lopez/Volume 1 (2013)
Though many of the intricate details of Field’s view have only been canvassed in the foregoing
discussion this notion of entitlement is interesting in the way it attempts to work within the
foundationalist framework. Ultimately, Field is going to say that when we get to the really basic beliefs
all we can do is express our attitudes toward these beliefs, give it a thumbs up or thumbs down. At least
with respect to classical foundationalism, the skeptic wants more than just evaluative attitudes to justify
basic beliefs as these wouldn’t have an indubitable status. The foundationalist, if interested in using
epistemic entitlements as foundations, will not find much solace on Fields view.
IV Quinean Holism and the Web of Beliefs
The foregoing discussion and subsequent evaluation of epistemic entitlements was intended to show that
they fall short of the mark of serving as foundations for our knowledge as the foundationalist might want.
In this section I hope to recast epistemic entitlement as playing a central role in a Quinean web of beliefs.
Similar to Duhem before him, Quine maintains that we cannot test a hypothesis in isolation
because auxiliary assumptions will always be necessary for a prediction to be tested/evaluated. Duhem
offered two thesis to explain the role of disconfirming/confirming evidence in scientific theories:
a theory does not entail an observation rather,
a theory in conjunction with auxiliary hypothesis entail an observation.15
In support of his theses Duhem offers that it is never the case that a theory can be tested in
isolation. Rather, an experiment in conjunction with background/auxiliary assumptions contribute to the
observation obtained. For example, a simple measurement of expanding gasses will presuppose the ideal
gas law and that all relevant apparatus are working appropriately. If one obtains unexpected evidence then
it is not the theory that has been falsified but rather any of the background assumptions or any of the
instruments could be responsible for the error. As such, it is ambiguous solely based on this experiment
what has been disconfirmed.
15
This notation is taken from J. A. Cover and Martin Curd, “Commentary,” in Philosophy of Science: The Central
Issues, eds. J. A. Cover and Martin Curd (New York: W.W. Norton & Company, 1998), 334.
52
Jonathan Lopez/Volume 1 (2013)
Quine’s Holistic Epistemology is much wider in scope. It maintains that all theories, not just
scientific ones, presuppose a set of background assumptions and are tested together rather than
individually. For example, if someone had a simple hypothesis such as “the bus will soon arrive at this
stop” which is confirmed by there being many people at the particular bus stop it is more than the
hypothesis that is confirmed. Quine would say that all the assumptions that fed into this hypothesis were
also confirmed, such as: the amount of people at a bus stop is directly proportional to how long the last
bus was at the stop; busses stop by this stop at regular intervals; people combine additively; etc.
When it comes to revising our beliefs Quine suggests we employ the maxim of minimum
mutilation, which roughly instructs us to make the most conservative of adjustments to a theory when
revising it. Say we have a set of purported truths, S, that make up a web of belief which have been found
to jointly imply some falsity.16 One or more of the sentences that constitute S will have of be rescinded
such that the false implication is removed. The members of S we should rescind are the ones that seem
most suspect, or least crucial to the overall theory. Repeating this process will eventually yield a set of
beliefs, S`, which will be free of the false implication.
It is important to point out here that Quine is not offering a Coherentist picture, but rather one that
is not strictly classical foundationalism. Quine does not distinguish between foundational/basic and
derived beliefs because everything is taken to be part of the same web and on the same level. For it to be a
Coherentist picture it, one would have to stipulate further that beliefs mutually support each other. The
only difference between beliefs that Quine recognizes is that some are more central to the web, the ones
hardly ever revised, in virtue of being connected to many different kinds of beliefs. The difference is that
an entire web of belief rests on all our experiential knowledge, so it can be thought of as foundational in
this sense. The Coherentist would say that some system of belief would support itself and in a sense
“float” not needing to rest on anything.
An example of something at the center of Quine’s web of beliefs is classical logic. Within
Quine’s web of beliefs there is nothing that is immune to revision though the maxim of minimum
mutilation does protect central beliefs. With the development of Quantum mechanics, however, some
have noticed that it may be expedient to amend classical logic, specifically reformulate the Law of
16
W. V. O. Quine, Pursuit of Truth (Cambridge: Harvard University Press, 1992), 14- 15.
53
Jonathan Lopez/Volume 1 (2013)
Excluded Middle.17 This would be perfectly acceptable for Quine given that the change does lead to a
better understanding overall. This issue will not be explored further it is simply mentioned to show that
nothing is off bounds for revision within a web of beliefs, there are no true foundations.
Quine and Wright seem to be referring to a very similar epistemic notion in speaking of central
web beliefs and “cornerstone’ propositions, respectively. Further, Wright’s epistemic entitlement notion
helps us see what is at the center of the web of belief. Wright’s cornerstone propositions basically
propped up an entire region of belief. That is, a cornerstone belief is a very important (although
pragmatically so) because doing without such would leave one unable to reason in that region of belief.
Quine identifies a belief as central to a web if it is intertwined to many other beliefs. Though a central
belief does not do any ‘propping up’, if doubted will lead to all the intertwined beliefs to being amended.
We privilege something like logic or sensory experience for very similar reasons. Quine says these will be
at the center of a web of belief while Wright says these are ‘cornerstone’ propositions and further we are
entitled to which are conducive to rational decision-making or cognitive processes. Essentially both are
saying that these are indispensable. Though Wright will go further and say that such a cornerstone is a
“free foundation” as we have seen he falls short of the mark of what a foundationalist might ask for in a
foundation because of his appeal to pragmatic rather than epistemic grounds.
V. Entitlements as Anchors
Though I do not take it as settled that foundations for our knowledge cannot be found I hope it is at least
seen that the Epistemic Entitlements explored here will not fit the bill. Falling short of this mark, I’ve
attempted to recast the Epistemic Entitlements discussed into a role wherein they can help us better
understand Quinean web of beliefs. I’ll now turn attention to further developing a non-foundational
picture with a humbler understanding of Epistemic Entitlements.
In the absence of foundational propositions to ground our knowledge Neurath offers the
following metaphor to describe a non-foundational picture of knowledge:
We are like sailors who on the open sea must reconstruct their ship but are never able to start
afresh from the bottom. Where a beam is taken away a new one must at once be put there, and for
17
Hilary Putnam, “Is Logic Empirical?,” Boston Studies in the Philosophy of Science 5 (1968): 216-241.
54
Jonathan Lopez/Volume 1 (2013)
this the rest of the ship is used as support. In this way, by using the old beams and driftwood the
ship can be shaped entirely anew, but only by gradual reconstruction.18
An important part of what Neurath is trying to capture in this metaphor is a non-foundationalist picture of
knowledge. Unlike Descartes who thought we were able to build our knowledge on firm foundations,
Neurath offers a much more modest picture. For Neurath, while one is able to build up knowledge it is not
on firm grounds (i.e. not foundations). Instead, in accordance with Quine, the only way we are able to
revise our knowledge is by “gradual reconstruction” never being able to take the whole thing apart.
Although strictly speaking all beliefs are tested in unison, for Quine, some are more central to a
web of belief than others. Wright also prioritizes certain beliefs over others, except one difference is that
‘cornerstone’ propositions will serve as foundations for our knowledge. Though Quine and Wright arrive
at why a belief is to be at the center of a web of belief or a ‘cornerstone’ proposition, respectively, for
different reasons a salient similarity is that both think these beliefs are integral to holding our beliefs
together. Where they certainly depart is the degree to which they think these beliefs, web centers or
cornerstones, affix our beliefs. For Quine the center of a web of belief is still part of the web and as such
some recalcitrant observation is fully capable of displacing it. Wright, in the foundationalist spirit, wants
to say that we can use cornerstones as foundations but as we have seen offers pragmatic grounds for this.
Falling short of the foundationalist goal of firm foundations, we can salvage Wright’s notion of epistemic
entitlement into a non-foundationalist picture. All the propositions that Wright was calling ‘cornerstones’
could alternatively be thought of as identifying the center of a web of belief. On this conception we would
have many loci in a web which departs a little from the original Quinean notion. Interestingly Wright’s
idea of having regions of thought is still captured nicely by this new picture though ultimately all regions
are to be intertwined.
The upshot of a web of belief with multiple loci, or anchors, is potentially helpful to the
coherentist. Among the difficulties with coherentism as a theory for justification is that it is possible for
anyone to create a coherent system of propositions that might be completely out of touch with reality, this
is sometimes called the isolation objection. For example, if I wanted to believe I was the King of Zembla
I might create a coherent system by inventing or distorting facts such that they are coherent with my
initial proposition. I might say to myself that I am in exile to explain my lack of power or interpret the
18
Nancy Cartwright, Jordi Cat, Lola Fleck, Thomas E. Uebel, Otto Neurath: Philosophy Between Science and
Politics (Cambridge University Press, 1996), 89- 94.
55
Jonathan Lopez/Volume 1 (2013)
works of famous poets as stories of Zembla. However, upon being asked where Zembla is exactly the best
I could do is gesture to some distant northern land. At this point I would no longer be respecting the
cornerstone proposition/anchor that my senses are putting me in touch with the external world since there
does not exist an observable geographic location known as Zembla. At this point I would have to realign
my beliefs making sure that this cornerstone/anchor is respected. The coherentist response to the isolation
objection would argue something along the lines that cornerstone propositions or anchors are what allows
a web of beliefs to become adequately compatible with external world that exists outside of their coherent
system, although they may want a more rigorous way to determine what is an appropriate anchor.
As the discussion in previous sections intended to show, epistemic entitlements are not
foundational and we should instead think of epistemic entitlement as ‘anchors’ on Neurath’s boat. They
are disqualified from the rank of foundations because at least on the classical picture they will not be
some indubitable basic belief upon which the skeptic can build knowledge. Neurath’s metaphor is helpful
in understanding how the epistemic entitlements can be adopted into a non-foundationalist picture by
giving them the name anchor. The thinking behind the recategorization is that falling short of being
foundations they still stabilize our corpus of knowledge because without them we certainly wouldn’t be
capable of many of the cognitive projects we undertake. In his metaphor, Neurath points out that when we
rebuild our knowledge, rebuild the ship, we can only use other parts of our knowledge, the other planks,
in this endeavor. But surely the process is not as haphazardly as the metaphor might suggest since there
are some beliefs we have more confidence we use grounds to jettison other beliefs, Quine captures this
nicely with his notion of web of beliefs. An anchor is supposed to fill this role, a sort of quasi-foundation,
we provisionally put more confidence in such that we can amend the rest of our beliefs. It’s important to
note that, similar to boats, anchors only stabilize they do not fix knowledge as a foundationalist seeks to
do with foundations.
56
Download