I have finished my thesis and I am happy to announce that I won the best presentation award from the Biola Philosophy Department. I was blown away by the award and it turns out that I am a very poor judge of my own presentation skills. The award included a book of titled The Basic Works of Aristotle edited by Richard Mckeon, and getting my name on the Best Presentation plaque in Sutherland Hall.
Here is my paper for anyone that might be interested in reading it.
Any questions, critiques, and thoughts welcome.
Feeling Isolated? You May Just Be Incoherent
A Critique of Coherentism and the Isolation Objection
By Kyle Dickey
Phil 450
Biola University
Introduction
The coherence theory of justification is a highly debated and discussed epistemological theory. It is an attempt to answer a number of basic questions about beliefs and how they become justified. It is a robust theory that has been defended by a number of prominent philosophers in the last century. In this paper I will discuss the coherence theory of justification and the most devastating issue raised for the theory: the isolation objection. I will then discuss Laurence BonJour's and Keith Lehrer's versions of the coherence theory which I consider the two best formulated versions of the theory. These two are significantly different from one another, but each has its own way by which to answer the isolation objection. And in closing I will raise my own thoughts to how combining these approaches might be useful in answering the objection more thoroughly.
Basic Coherence
Many epistemologists have posited the existence of basic beliefs in order to answer the skeptical regress argument. This foundationalist answer to this problem claims the existence of basic beliefs stops the regress of justification by not being justified by other beliefs. Basic beliefs are thought to be self-justified or justified by something other than beliefs like sense experiences. These basic beliefs act as foundations for the larger structure of beliefs. Coherentism rejects this theory of basic beliefs, claiming that justification cannot be conferred on a belief by anything but other beliefs. Thus the coherentist rejects the foundationalist distinction between basic and regular beliefs, giving in its place the notion of coherence as the primary source of justification. Simply put, the justification of beliefs in a coherentist system is due to the relations between beliefs and the level to which they generally cohere together. In contrast to foundationalist views, a coherence theory does not appeal to anything outside of the system of beliefs for justification.
The Isolation Objection
A number of objections have been lodged against the coherentist view of justification, the most devastating of which is the isolation objection. This objection generally claims that if a coherent system of beliefs is justified internally then it seems to be isolated from the external world. If such isolation occurs then coherence cannot be a valid way for one to arrive at truth. And because epistemic theories of justification are supposed to be attempts at getting to truth rather than falsehood, coherentism fails to be a tenable theory of justification.
There are three distinguishable types of objections called the isolation objection in the philosophical literature; it will be helpful to differentiate between them. The first type of isolation objection is called the alternative coherent systems objection. It says that a coherence theory of justification claims any system of beliefs constituting knowledge can only be justified by internal coherence. This makes it impossible for there to be one uniquely justified system of beliefs simply by appealing to coherence. If for a system of beliefs to be justified it need only to be coherent, then any number of systems (if not an infinite number) can be found as equally coherent yet incompatible. Here there can be no non-arbitrary choice between any of these systems on the basis of coherence. Thus each belief can be equally justified which results in skepticism about truth. One system of beliefs may be justified including belief X while another is equally justified including belief ~X. We are no closer to the truth because both X and ~X can be justified, but justification seems aimed at getting to truth. As a result coherence can only commit one to skepticism about the truth of beliefs.i
The second type of isolation objection, the input objection, states that coherence is only a matter of internal relations between the beliefs in a belief system. There is no connection or reliance on anything external to the system. On a coherence theory an entire system of beliefs might be justified and even constitute knowledge, but nothing about coherence causes beliefs to be in contact with the outside world they supposedly describe. Coherence among beliefs does not require that there be any input from the world that causally influences a system of beliefs. Thus if the beliefs in a coherent system were to match up with the world it could only be accidental. The accidental truth of beliefs cannot constitute knowledge or contribute to justification.ii
The third type of isolation objection is the problem of truth. A basic part of all epistemological theories of justification is that they must show themselves to be truth-conducive. The objection states that a coherence theory of justification cannot accomplish this task without also committing any proponents to a coherence theory of truth and an idealist metaphysics. These theories taken together create the only way coherentist justification can match up with truth. However, there is much debate as to whether these theories are correct when it comes to truth. These two theories are often seen as extremely unsatisfactory and untenable. Thus any theory that commits one to other untenable theories is itself highly unsatisfactory.iii
A number of coherence theorists that have responded to each of these objections and the isolation objection as a whole. Most of these responses have come as part of the formulation of different coherence theories. Two of the most prominent proponents of the coherence theory are Laurence BonJour (until recently) and Keith Lehrer. I will now lay out the basic components of their theories and the ways in which these theories deal with the isolation objection.
BonJour's Coherence Theory
Of all the recent and contemporary coherence theories to be propounded, BonJour's theory has been the most influential and discussed. Although BonJour no longer holds to his theory, having converted to a type of foundationalism, his coherence theory seems to come the closest to a classic coherence theory while also accounting for the isolation objection. BonJour's basic theory can be divided into four parts: (i.) non-linear justification, (ii.) the concept of coherence, (iii.) the observation requirement, and (iv.) the doxastic presumption.iv
(i.)Non-linear Justification
One distinction that must be made in order to lend plausibility to a coherence theory of justification is the difference between linear and non-linear justification. On a foundational view a justified belief is directly justified by another belief. In this way justification happens in a sort of line where one belief is justified by its prior, and that belief is justified by its prior, etc. Coherentists see this kind of justification as problematic. If justification is said to function in a kind of line and a coherence theory of justification is true then there would be a viscous regress of justification. One belief would be justified by a prior belief which in turn would also be justified by a prior belief. Eventually the original justified belief would have to be appealed to as a justifying premise for a belief far below it on this justification chain. Thus a circle would be made causing a belief's justification to ultimately rely upon the itself. This is an unacceptable result, and as mentioned earlier, the coherentist denies the existence of basic beliefs which foundationalists appeal to in order to solve this problem. So the coherentist must claim justification to function in a different sort of way.
BonJour appeals to the idea that justification when properly understood should actually be seen as non-linear or holistic in nature. This kind of view, first formulated by Bosanquetv, avoids both the asymmetry and circularity of a linear theory. It does so by claiming the true source of justification to be the relevant system of beliefs and the relations of mutual support between them. A belief is justified by its relation to the believer's entire justified system of beliefs, not by inference from prior beliefs as on a linear model.vi
(ii.)The Concept of Coherence
For BonJour the concept of coherence is still a widely unclear idea. The topic is broad and deep in philosophy but has scarcely been well formulated. There are however a few relatively clear ideas which constitute the basic make-up of coherence. The parts that he gives as being necessary for coherence are logical consistency, a high degree of inferential connectedness, probabilistic consistency, and explanatory relations between beliefs.
Logical consistency is the most obvious factor in coherence. This condition is such that no beliefs in one's system of beliefs logically contradict one another. A system of beliefs would be incoherent if it included both a belief and its negation. But logical consistency cannot be the whole story of coherence. It is easy to think of a system of beliefs that is highly consistent with no contradictions and yet none of its beliefs relate to each other. It would seem very odd to call such a system coherent simply because none of its parts are in opposition. Thus a high degree of inferential connectedness is also a main ingredient in the concept of coherence.
The more coherent a belief system is on a whole, the more interconnected and related different beliefs and relevant systems of beliefs are. The more support and evidence beliefs can lend each other, the more coherent and justified the system is as a whole. Probabilistic consistency is a sort of sub-requirement under inferential interconnectedness for BonJour. The coherence of a system is enhanced when a belief lends a probability of truth to other beliefs. Thus the more probable it is for one belief to be true, the probability of the beliefs it has connection and relation to also increase.
Explanatory relations also play a role in the coherence of beliefs. The ideal form of coherentism would be some sort of unified system much like the attempts at a unified theory in science and physics. In such a system all beliefs would cohere together and be explained in one simple and unified fashion. While this is extremely desirable, BonJour seems to think it improbable and problematic to hold such a view because no such theory or system has been developed without falling prey to serious objections. However, the basic idea remains important to coherence theories. The more beliefs a system has lending explanatory power to other beliefs, the higher the degree of coherence in the system. Seeking out explanatory relationships and giving background evidence is a highly desirable feature of a coherence theory. BonJour believes none of these requirements meet the necessary and sufficient conditions of what coherence ought to be, but all are important parts of what such a theory should include.vii
(iii.)The Observation Requirement
At this point of BonJour's theory he begins to try and show how his theory might survive the isolation objection. He takes the second form of the objection, the input objection, to be the most serious and warranting of response. He establishes the observation requirement in an attempt to pacify the objection.
One part of a foundationalist theory that turns out to be highly desirable on most any theory of justification is the role it gives to sense experience. It intuitively seems very important that our sensual experience of the world and external realities play an important role in our beliefs. Foundationalism gives this experience a directly justifying relation to basic beliefs. Coherentism cannot claim experience as having any justificatory power because coherence theories claim only beliefs are able to justify other beliefs. Thus coherence theories of justification seem problematic unless they can allot for the importance of sense experience without also allowing it justificatory power.
BonJour attempts to solve this problem by introducing the notion of cognitively spontaneous beliefs (which I will hereby call C.S. beliefs). These he claims are empirical beliefs that are non-inferentially arrived at as a causal result of sensory experience. Because they are not inferred from anything, sensory experience cannot be appealed to as providing justification for these beliefs. The idea is that C.S. beliefs are a kind of belief that simply strikes an observer in an “involuntary and coercive way.”viii An example might be that I believe there is a computer in front of me. I simply find myself holding such a belief. These observational beliefs have a special kind of status. They have been arrived at by being caused through sense experience, but this is not a factor in their justification. These beliefs are justified within a person's system of beliefs by coherence with a number of other beliefs. These other beliefs might pertain to the spontaneous occurrence, relevant background conditions, and the apparently successful track record of truth for C.S. beliefs under the relevant conditions. All conditions being assessed, a cognitively spontaneous belief that coheres with a belief system can be seen as justified. Such C.S. beliefs fulfill both the intuitive role that sensory experience must play in beliefs and the coherence requirement that justification be solely a matter of relations between beliefs.
The observation requirement for BonJour must be seen as a necessary requirement. If observation is claimed only to be a possibility for a coherence theory BonJour thinks the theory to have failed. The intuition that many have of sense experience and the role it plays in relation to beliefs seems necessary. Sense experience seems to play a very important role in most of the beliefs we have, especially those of empirical nature. Thus a theory that accommodates these intuitions must contain a significant requirement for observation.
Lastly, while still discussing observation, it is important to look at what justifies these C.S. beliefs. As stated above it is the coherence that C.S. beliefs have with beliefs about the fact of spontaneous occurrence, the relevant background conditions, and the track record of truth with respect to frequency under such conditions. These beliefs themselves must also be justified, and while it seems that other C.S. beliefs might well be appealed to for their justification, there is an element of introspection that must also be addressed. This is especially pressing when deciphering the justification of beliefs about the frequency of correctness C.S. beliefs have under the relevant conditions. An appeal to introspective and meta-beliefs must be made, and this leads us into BonJour's fourth and final coherence condition.ix
(iv.)The Doxastic Presumption
The final condition of BonJour's theory of coherence is an appeal to what he calls the doxastic presumption. He needs this condition to satisfy a problem that his coherence theory seems to run up against when it comes to the justification of internal introspective beliefs and meta-beliefs. The problem is that from an internalist point of view beliefs about the relevant conditions lending justification to empirical beliefs need to be justified themselves. This justification seems only possible in appeal to meta-beliefs about the truth of the relevant condition beliefs. These meta-beliefs would constitute a sort of reflective grasp of the fact that one's system of beliefs satisfies the observation requirement, and would give the justifying basis for one to hold to such a system as a whole.x But if these meta-beliefs are to be justified by appeal to coherence then an unavoidable vicious circle would be created. In such a circle the meta-beliefs would be appealed to in order to justify beliefs about the conditions of coherence and coherence would be the justifying factor of these meta beliefs. What then is it that justifies these meta-beliefs? There is no justifying factor, replies BonJour. Only what he takes to be an appeal to the doxastic presumption.
What he means by the doxastic presumption is to treat these meta-beliefs as an “unjustified hypothesis in relation to which issues of justification are conditionally assessed, yielding results of the general form: if my representation of my whole body of beliefs is correct, then such and such particular belief is justified in the sense of being likely to be true.”xi The idea is that if one presumes that the representation of their own system of beliefs is likely to be true, then beliefs about the conditions of coherence can be justified and coherence can act as the justifying factor for all other beliefs one may have.xii
BonJour's theory takes these four conditions to be the basics of what a coherence theory should include. Each condition takes on a number of issues that have plagued coherence theories of justification and gives cogent answers and plausible lines of response to these problems.
Keith Lehrer's Coherence Theory
Keith Lehrer's coherence theory of justification from his book Theory of Knowledge is another highly discussed version of what coherence might amount to. The main ingredients of his theory include (i.) the Acceptance System, (ii.) Competition and the Justification Game, (iii.) the Principle of Trustworthiness, and (iv.) Verific Justification.
(i.)The Acceptance System
The acceptance system for Lehrer is just the things that one accepts in the “interest of obtaining truth and avoiding error.” Coherence within an acceptance system depends on what it is reasonable to accept based on the system. This system is meant to provide one with personal justification and thus provides a subjective form of justification internally available to the person.
First, it will be helpful to understand that Lehrer uses the notion of acceptance instead of belief. He sees acceptance as a more useful and functional idea and often leaves out the common notion of belief when talking about justification. He does this in order that he can claim the acceptance system to be the central component in one's attempt at obtaining truth and avoiding error. Beliefs for Lehrer can oftentimes be wrong, and we often do not accept our own beliefs (perceptual or other) because we have evidence contrary to their truth.xiii
The acceptance system provides justification by telling us what it is more reasonable to accept “when we seek truth.”xiv The system determines the reasonableness of differing or competing acceptances on the basis of background information. Sources, circumstances, and undermining objections all play a role as information in this system and the more trustworthy this information is the more reasonable the acceptance backed by the information becomes. Lehrer takes the notion of reasonableness as primitive in order that he need not define the relevant factors deciding what it is to reasonably accept one thing over another. This is a highly contested issue he leaves to be addressed by others.xv
(ii.)Competition and the Justification Game
Competition is the the process by which the acceptance system comes to hold one claim over another. Two claims will often compete in this way. A claim and its negation may be deliberated over in deciding which is more reasonable given one's acceptance system. Or an even more difficult verdict may be drawn between a claim (say a perceptual claim) and one of skeptical nature (e.g. that one is hallucinating). The acceptance system decides which claim is more reasonable on the basis of this competition. A claim wins out in this competition by beating all of its competitors. Thus a claim that beats its competitors is the more reasonable claim to hold. By this reasonableness, provided by the acceptance system, one is personally justified in holding the claim.
Deciding how a certain claim is beaten in this competition brings up what Lehrer calls the justification game. He uses this justification game as an example to portray the competition between claims in the acceptance system. The game begins when one puts forth a claim one accepts to be true. A skeptic may then respond with any objection in the form of a competing claim. The original claimant can only beat the competing claim if it is more reasonable for the claimant to hold to his original claim than the skeptical claim. If the claimant beats the competitor than he wins the round, and if he beats all the competitors the skeptic raises, he wins the game. Lehrer claims this example to parallel with the way the acceptance system works. Coherence plays the role of personal justification in how well one's claims cohere with one's acceptance system. The more trustworthy information one has backing up a claim the more reasonable a claim is within one's acceptance system. And the more trustworthy information or reasons one has backing up a claim the better it coheres with ones acceptance system. In this way coherence confers personal justification on Lehrer's theory.
At this point a slight addition must be made to the notion of competition. Lehrer believes another way a skeptical objection might be overcome is by being neutralized. Often skeptical competitors come in a form that is unrelated to the claim at hand. Such a case might be one's claim to see a red book in front of them and the skeptic replies that people sometimes see red books while dreaming. In such a case Lehrer appeals to a conjunctive statement as neutralizing this competitor. This might be the claim that people do sometimes see red books while dreaming and that I am currently not dreaming. If one has no reason to believe that one is dreaming it seems to be more reasonable to accept this conjunctive statement then the skeptical competitor and the competitor is neutralized. If a competitive claim is beaten or neutralized a person may be personally justified in their original claim.xvi
(iii.)The Principle of Trustworthiness
At this point Lehrer's theory relies heavily on the idea that one has trustworthy information, and that one's acceptance system is a trustworthy guide to truth. To gain such information one must believe themselves to be trustworthy as well as trust their faculties (perception, memory, reasoning, etc.). Thus much of Lehrer's theory relies on a principle of trustworthiness which he introduces:
“T. Whatever I accept with the objective of accepting something just in case it is true, I accept in a trustworthy manner.”
If one is to accept this principle on the basis of coherence with the acceptance system one's trustworthy information can become personally justified. But Lehrer does not say that one has to be personally justified in accepting the principle itself. Only that one needs to accept the principle as more reasonable than its denial. Thus on the basis of the reasonableness of one's trustworthiness and the coherence of one's claims with their acceptance system, any competitors to the claims one accepts can be beaten or neutralized. In this way one is personally justified in holding a claim.xvii
(iv.)Verific Justification
The final part of Lehrer's theory he calls verific justification and he appeals to it as part of a person's ultra system. The ultra system is the combination of both a one's acceptance system and their verific system. This combination system answers the intuitive worry that though someone may be personally justified they still might be wrong and not completely justified. The verific system is just a subsystem of the acceptance system where all of one's acceptances turning out to be false are simply deleted. The idea here is that complete justification is personal justification that is not based on error. One can only be completely justified in a case where a claim is a member of one's ultra system which includes their acceptance system and verific system. It seems that only an omniscient observer would be able to verify the members of ones acceptance system worthy of verific justification, but this is not a problem for Lehrer. His goal is that his coherence theory gets the correct answer in situations where a person may be justified and yet their acceptance is false. In competitive situations where one is ignorant to the truth or the reasonableness of a skeptical worry, one may be personally justified in holding to their claim. Yet, in the ultra system they would fail to be completely justified because they lack the truth necessary for verific justification.
Verific justification and the members of an ultra system are not accessible to one's subjective states in the same way as the acceptance system. Thus for Lehrer the best one can do is try their hardest to avoid error in what they accept. In our search for truth there are definitely a number of things we are personally justified in accepting. Thus whether or not our personal justification relies on an unseen error we can win the justification game. Yet if error is present Lehrer believes we have “lost the prize of knowledge.”xviii
Dealing with the Isolation Objection
Now with both BonJour's and Lehrer's theories laid out it can be made clear how each deals with the isolation objection. Each answers this problem in sufficiently distinct ways and both give substantial replies responding to differing intuitions. I will lay out these replies and some objections to each. I will then give my own thoughts on how combining BonJour's and Lehrer's replies to these objections might be a better way of answering intuitions behind the isolation objection as a whole.
BonJour and the Observation Requirement
BonJour takes the the input objection to be the most pressing form of the isolation objection. He believes a sufficient answer to this problem pacifies the other objections. He does this by introducing the observation requirement into his coherence theory as I have laid out above. This answers the input objection by making it a necessary requirement of his theory that observation play a central role in one's belief forming process, and that one's beliefs about sense experience are justified through their coherence with the larger relevant system of beliefs. The intuition behind the input objection is that sense experience plays a largely important role in our system of beliefs, and it seems that BonJour answers this intuition thoroughly. He posits the existence of C.S. beliefs which gives sense perception a necessary role in his coherence theory. In this way sensory input acts as the causal factor for C.S. beliefs and C.S. beliefs are required for a belief system to be coherent. By requiring input into his coherence theory through the observation requirement BonJour answers the worries of the input objection.
Many objectors such as Goldman, Silvers, and Kornblinth raise questions to BonJour as to whether the observation requirement is necessary for a coherence theory of justification.xix Lehrer himself denies the necessity of any observation requirement claiming that its incorporation into a coherence theory is either unneeded or arbitrary. The common objections of this sort claim that if our system of beliefs give us no justification for our observation beliefs then the requirement is arbitrary, but if the system of beliefs does give justification to our observational beliefs then the requirement is unnecessary.xx I disagree with both horns of this piercing dilemma.
It cannot be such that the observation requirement is arbitrary or unnecessary. The reason behind incorporating the observation requirement into a coherence theory of justification is in order to answer the input objection. If there were no requirement for input into a coherent system then the input objection would would succeed in showing coherence theories to be isolated. Thus the observation requirement is a necessary part of responding to the input objection and cannot be seen as arbitrary. If a system of beliefs, on a coherence theory including an observation requirement, fails to provide justification for observational beliefs, a lack of input cannot be claimed at fault. If a system of beliefs does provide justification for observational beliefs on a coherence theory without an observation requirement there would seem to be a lack of the necessary connection between input and coherence.xxi Thus it seems much more desirable to incorporate an observation requirement into a coherence theory in order to quench the input objection, making it an essential and non-arbitrary part of such a theory.
The observation requirement is a necessary part of a coherence theory because it answers the underlying intuition of the input objection. This intuition is that sensory experience plays a significant role in our belief formation and justification. Thus the observation requirement makes it a necessary part of the coherence theory that sensory experience play a central role in belief formation, while also appealing to coherence as the justifying factor for these beliefs. It seems at least some sort of observation requirement is necessary for answering the intuitions of the input objection.
BonJour believes that his appeal to an observation requirement sufficiently solves the input objection while also getting at answers to the other forms of the isolation objection. He claims that answers to both the alternative coherent systems objection and the problem of truth are easily found once a coherence theory requires sensory input into a belief system.xxii Both objections rely on intuitions that coherence theories have problematic truth connections and once input from the external world is seen as a truth connection for coherence theories the objections become unproblematic. He delves deeper into more specific answers to each objection, and though I agree with much of what he says, for my purposes here I will not go into further detail. Instead I will look at an interesting argument of Lehrer's and the way in which his theory provides answers for these objections.
Lehrer's Transformation Argument
Lehrer's theory provides a very useful argument against the isolation objection that seems extremely helpful for a coherence theory. He uses his laid out theories of the acceptance system and competition within the justification game to show that any arguments concluding coherence theories to be isolated from truth or reality can be beaten or neutralized. He calls this move the transformation argument and it appeals to his idea that any claim made by a skeptic or in this case an isolation objector must enter into one's acceptance system as a competitor. Thus a claim, such as “you are isolated from the external world,” must be seen as a competitive claim to other claims one may have about being in contact with the external world.xxiii One's acceptance system then becomes the deciding factor between the two claims. If it turns out to be more reasonable to accept a claim that one is in fact connected to the external world (if the claim coheres better with their acceptance system), then one is personally justified in holding to that claim. If the objector's claim turns out to be less reasonable to accept then it is beaten or neutralized and not personally justified for the claimant.
In the ultra system and the verific justification game if the claim that one is not isolated from the external world turns out to be true then the person is completely justified. The undefeated justification of this claim transforms it into knowledge. If on the other hand the claim fails to be true and the objector's skeptical worry of isolation is correct, then the coherence theory would still give the right result. The claimant in question might be personally justified in holding to their claim but not verifically justified. Thus the claim would not be a member of their ultra system because it would fail to be completely justified. This claim would lose in the justification game because its competitor would not be beaten or neutralized.
Lehrer's theory can thus avoid any skeptical worries the general isolation objection might present. Any claim that is completely justified is a part of the ultra system. And any claim in the ultra system is both personally and verifically justified. For the personal justification of one's claim to remain undefeated, “it must be true that one is connected in the way one accepts that one is.” This truth connection transforms personal justification into knowledge.xxiv
This transformation argument of Lehrer's is a highly useful tool in answering the isolation objection. It provides an interesting answer to many problems a coherence theory might have in connecting up with truth and reality. In reply to the general isolation objection Lehrer's theory can appeal to verific justification and come to the correct conclusion. Lehrer's response to a number of his objector's has used this transformation argument in just such a way. Many papers in response to Lehrer's theory such as those of Moser, Mattey, and Russel have tried to critique Lehrer's position by giving examples where one is personally justified and possibly correct in their beliefs or acceptances but fail to gain knowledge.xxv
In reply to each objector Lehrer uses the transformation argument to show that his theory accommodates for each of their intuitions. In each case, the objection that one does not have knowledge or is isolated from the external world must be seen as a competitor. Thus, in order for one to be personally justified in holding their original claim they must have trustworthy information reasonable enough to beat or neutralize this competitor. If it turns out they really are personally justified and still do not attain knowledge it must be due to an error which will be caught in the verific system and one will fail to be completely justified. If it turns out that their acceptance is true but they are not justified in the correct way then they fail to attain personal justification. Thus the answer to any of these types of objections can be accommodated by the transformation argument.xxvi
The alternative coherent systems version of the isolation objection can be answered in a similar way. The transformation argument, under Lehrer's ultra system, is able to claim only a single set of acceptances turn out to be true. The argument allows for there to be a number of incompatible coherent acceptance systems but only one comprehensive set of acceptances that can be verifically justified. In this way the argument can answer both the intuition that sometimes a person may be justified but not correct and the intuition that there must be only one set of true beliefs or acceptances that match up with reality.
The transformation argument also allows for a sufficient answer to the Problem of Truth. Part of Lehrer's main objective in laying out his epistemic theory is getting at the truth. He claims that what one accepts is done so “in the interest of obtaining truth and avoiding error” and uses the acceptance system as getting at truth in this way. Thus, on Lehrer's coherence theory of justification, if a coherence theory of truth or idealist metaphysics really are untenable theories one would fail to be completely justified in holding them. If one lacks the trustworthy information making it reasonable to hold to these theories then one is not personally justified in holding to them. On the other hand if one is personally justified in holding these theories but they turn out to be incorrect then the one holding the theories would fail to be completely justified. Thus, once again Lehrer's coherence theory can correctly answers our epistemic intuitions. If one holds to both a coherence theory of truth and an idealist metaphysics and the theories turn out to be wrong then the verific system catches the error and the person fails to be completely justified. However, if the person holding these theories is personally justified in accepting them and they turn out to be true, then the person is completely justified and the competitors provided by the problem of truth objection are beaten or neutralized. Thus, undefeated justification allows for knowledge and the transformation argument is able to pacify any problematic intuitions an objector using this argument might find convincing.
Only the second type of the isolation objection, the input objection, seems to cause a problem for Lehrer's theory. The theory fails to accommodate the intuitions behind the input objection. As shown above, this objection presupposes an idea that many of our beliefs, specifically empirical beliefs, must be based on input from the external world in order to be true. Lehrer claims that his theory allows for the possibility of input but does not require it. A problem here for Lehrer's theory, similar to a critique that BonJour has made, is that for Lehrer a coherent system of beliefs or acceptances would be equally justified whether or not there is any external input.xxvii
On Lehrer's theory there is no requirement for observation. A claim may be completely justified for a person (personally and verifically justified) without any external input. This seems contrary to the intuitions that base the input objection. If a person is to be completely justified in a claim about the external world without any external input, then it would seem only possible for the claim to be accidentally true within the ultra system.xxviii This accidental truth would not constitute knowledge.
Lehrer fails here to accommodate the intuitions of objectors and even other coherence theorists (those of whom include observation requirements) that if one is to have a justified empirical belief it must be based on some sort of external input. Lehrer's theory does allow for such input to be possible but doesn't place it as a requirement. Without accounting for or including a requirement of observation into his theory Lehrer opens it up to this possibility of knowledge being undefeated accidental justification. Such a justified belief would go undefeated in the ultra system but would only be true accidentally. There would be no input from which the belief would be caused but it would nonetheless be true. Because Lehrer claims knowledge to be undefeated justification in the ultra system his theory would also allow such accidental true belief to be knowledge. But this is an unacceptable result on any justification theory.
The transformation argument cannot help Lehrer's theory here either. An objector might propose a claim that one's coherence theory relies on nothing external to it, or that one's beliefs have no external input. These claims would act as competitors to the claims that one's theory does rely on something external or that one's beliefs do have external input. But it seems without some kind of background observation claim, one could not be personally justified in holding either of these further claims. One would have no trustworthy reason able to beat or neutralize the input objector's competitors.
If one were to have a background observation claim that could beat or neutralize the input objection and cohere with one's acceptance system, then one would be personally justified. It is hard to see a way one might have such a claim without appealing to an observation requirement. It seems that for any claim in an acceptance system of an empirical nature, there would also necessarily be a background claim relating to observation and explaining how the empirical claim was arrived at. If this condition is not met then the acceptance would never be able to constitute a piece of knowledge. The claim would either fail to be reasonable enough to be in the acceptance system or fall into this category of undefeated accidental truth. So it seems that for any empirical claim to become knowledge there must be a background claim of observation explaining the claim and lending to its reasonableness. This it seems is an observation requirement.
Lehrer's theory then seems to need an observation requirement to deal with this problem of accidental truth. If his theory was to incorporate such a requirement it would have a better chance at achieving knowledge of empirical beliefs. Also, the transformation argument would be bolstered in such a way as to be able to answer the input objection, thus giving it the capability to provide an answer to each type of isolation objection laid out above. I conclude that, on a scale of attractiveness, Lehrer's theory would be more desirable if it was to include an observation requirement.
Conclusions
To this point I have argued in three main directions. First, that an observation requirement, like that of BonJour's, is not an arbitrary or unnecessary part of a coherence theory. Second, the transformation argument of Lehrer's theory provides an interesting and helpful way of answering most of the isolation objection. And third, the transformation argument falters in answering the input objection but could be bolstered if it included an observation requirement. At this point I wish to make a suggestion. It seems that a sort of combination approach between Lehrer's and BonJour's theories would be a strong and attractive form of coherentism.
I am not going to speculate on the exact ways in which such a theory might be arrived at, but I have shown that minimally the observation requirement and transformation argument are two useful bits of theory that seem to work well together in answering the most discouraging problems for coherence theories. BonJour and Lehrer have developed the most thorough theories explaining these two arguments and so I have used their theories to parse out my intuition that they might work well together. Thus a more in depth analysis into how these theories might be combined I leave for later consideration.
Relevance to Christianity
In my study of coherence theories of justification it has become apparent that foundationalism is the prominent epistemological view among Christians. This seems to be the case because Christians in the philosophical tradition are urgent in their search for truth and knowledge. Foundationalism has often been seen as the most plausible theory in arriving at this epistemic goal. Thus Christian thought has often been done in line with foundationalism. I would like to break that stereotype.
As I have laid out above the coherence theories of Lehrer and BonJour have at least a ring of plausibility to them. One I believe that foundational theories have had for a long time. It seems to however that coherence theories allow for a strong connection to be made with truth and maybe Christian epistemologists needs to look more closely at this connection. In all epistemic responsibility, if a Christian is truly looking for truth in epistemology then any and all widely held theories must be thought through and addressed. Coherence theories of justification should be looked at and thought about by Christians with all the rigor and strength that they often bring to doing religious philosophy. Maybe then Christian thought might grow within a plurality of epistemological views.xxix
(sorry, my end notes did not transfer over)
Bibliography
Audi, Robert. Belief, Justification, and Knowledge: an Introduction to Epistemology. Belmont, Calif.: Wadsworth Pub., 1988. Print.
Baergen, Ralph. Contemporary Epistemology. Fort Worth: Harcourt Brace College, 1995. Print.
Bender, John W. The Current State of the Coherence Theory: Critical Essays on the Epistemic Theories of Keith Lehrer and Laurence BonJour, with Replies. Dordrecht: Kluwer Academic, 1989. Print.
BonJour, Laurence, and Ernest Sosa. Epistemic Justification: Internalism Vs. Externalism, Foundations Vs. Virtues. Malden, MA: Blackwell Pub., 2003. Print.
BonJour, Laurence. The Structure of Empirical Knowledge. Cambridge, Mass.: Harvard UP, 1985. Print.
Bosanquet, Bernard. Implication and Linear Inference. New York: Kraus Reprint, 1968. Print.
Dancy, Jonathan. "A Defence of Coherentism." 1985. Comp. Louis P. Pojman. The Theory of Knowledge. 3rd ed. Belmont, Calif.: Wadsworth Pub., 2003. 206-15. Print.
Feldman, Richard. Epistemology. Upper Saddle River, N.J.: Prentice Hall, 2003. Print.
Fumerton, Richard. "A Critique of Coherentism." Trans. Louis P. Pojman. The Theory of Knowledge. 3rd ed. Belmont, Calif.: Wadsworth Pub., 2003. 215-21. Print.
Goldman, Alan H. "Bonjour's Coherentism." The Current State of the Coherence Theory 44th ser. (1989): 125-33. Print.
Haack, Susan. "Double-Aspect Foundherentism: A New Theory of Empirical Justification." Philosophy and Phenomenological Research LIII.1 (1993): 113-28. Print.
Kornblith, Hilary. "The Unattainability of Coherence." The Current State of the Coherence Theory 44th ser. (1989): 207-14. Print.
Lehrer, Keith. "Chisolm on Perceptual Knowledge: Foundationalism versus Coherentism." Metaphilosophy 34.5 (2003): 543-52. Print.
Lehrer, Keith. Theory of Knowledge. Boulder: Westview, 1990. Print.
Mattey, G. J. "Personal Coherence, Objectivity, and Reliability." The Current State of the Coherence Theory 44th ser. (1989): 38-51. Print.
Moser, Paul K. "Lehrer's Coheretism and the Isolation Objection." The Current State of the Coherence Theory 44th ser. (1989): 29-37. Print.
Sellars, Wilfrid, Richard Rorty, and Robert Brandom. Empiricism and the Philosophy of Mind. Cambridge, Mass.: Harvard UP, 1997. Print.
Russell, Bruce. "When Can What You Don't Know Hurt You?" The Current State of the Coherence Theory 44th ser. (1989): 94-102. Print.
Silvers, Stuart. "Coherence, Observation, and the Justification of Empirical Beliefs." The Current State of the Coherence Theory 44th ser. (1989): 168-77. Print.
Sosa, Ernest. "The Raft and the Pyramid: Coherence Versus Foundations in the Theory of Knowldge." Midwest Studies in Philosophy 5 (1980): 3-25. Print.
Stern, Robert. "Coherence as a Test for Truth." Philosophy and Phenomenological Research LXIX.No. 2 (2004): 296-326. Print.
