SEAT OF WISDOM Lecture Notes on Epistemology

SEMINARY, OWERRI-
. IMO STATE NIGERIA
Lecturer: Fr. Michael Nnamdi KONYE
Class: Philosophy Four Seminarians
2023/2024 Formation Year
Number of seminarians: 40
October/November 2023 Duration of lecture: 90 minutes

TRACT TWO: NORMATIVE EPISTEMOLOGY – Knowledge as Justified True Belief

Introduction:
The normative approach to knowledge attempts a comprehensive analysis of the conditions for
knowledge. These conditions have been traditionally listed as: belief (πιςτις -pistis -a mental state),
truth (αληθεια – aletheia - a judgement), and justification (λογος - logos - an evidence).
Since Edmund Gettier’s 1963 article which challenged these traditional three conditions (i.e.
justified, true belief) for knowledge, a renewed interest has been engendered within the academia, on
the viability of the normative approach to provide an adequate response to skepticism.
Critics, particularly from Willard von Orman Quine’s naturalized epistemology, have argued
for the replacement of traditional normative approach to epistemology but what would the search for
knowledge look like if stripped of such core values as justification and truth?
Today more than ever before, it becomes expedient to reaffirm those norms and criteria which
make for the understanding of knowledge as ‘episteme’ -i.e. stable, true, and reliable judgement about
knowable reality. For what it is worth, these basic criteria (JTB) approximate closer to what can be
called knowledge, since a knowledge claim is one that is justified irrespective of the required certainty
of a knower’s belief.

The value of propositional statements

To allow for a proper analysis of our knowledge claims, epistemologists would want to focus
on propositional statements which declare our belief states and thus represent purportedly our claims
to knowledge. How then do we identify which statements are said to be propositional? To begin with,
it is possible to distinguish kinds of ‘propositions’ or kinds of ‘statements’ (or ‘sentences’ as some
logicians may want to say) we can consider these propositional statements of our knowledge claims as
declarative statements, that is – those statements that are introduced by the demonstrative pronoun
‘that’. So, we can in fact speak of propositional and non-propositional kinds of statements.
For our purposes here, it might be helpful to distinguish apprehensive propositions which
simply describe states of affairs and does not make any commitment to truth or falsity from judicative
propositions which we simply refer to as judgements for the fact that they demand a commitment to
alethic values of truth and falsity. The focus of most epistemological considerations is on the judicative
propositional knowledge. This is so because truth is the goal of epistemological enquiry and as such it
is the normative object of the cognitive process at its peak – i.e. intellectual cognition or simply of
knowledge.
The other kinds of statements which are not ‘truth-laden’ can thus be called non-propositional.
In that same way, we also speak of propositional and non-propositional knowledge. Accordingly, we
can refer to propositional knowledge as a ‘knowledge that’, and a non-propositional knowledge as a
‘knowledge of’.
Nonetheless, all epistemic (epistemologically relevant) propositions are truth-laden. This is
because truth is the object of the knowing process, we can define epistemologically relevant types of

1
 propositions variously as: ‘the primary bearer of truth-values (true or false)’, ‘the (intentional) object
of belief or other propositional attitudes’, ‘the referent of that-clauses’, or ‘the meaning of a declarative
sentence’. It is in this sense that we usually speak of conditions/criteria for (true) knowledge.

Doxographical panorama (opinion poll) of the traditional conditions for knowledge.

The standard agreement since Plato’s Theatetus (cf. 201d – 201e) is that there are three
necessary conditions for any claim to knowledge: belief, truth, and justification. This is the normative
analysis of the components of knowledge so to speak.
Plato himself claimed that there are three ways of understanding knowledge: namely (1)
Knowledge as perception. (2) Knowledge as true belief. (3) Knowledge as true belief plus logos. Plato
favoured the last one (i.e. JTB) in which knowledge was seen as true belief combined with logos
(which last condition is today described as justification or warrant).
So also, in the Meno 98, he equally saw knowledge as true belief with right reason. The classical
definition for Plato then became knowledge as justified true belief.1 Contemporary philosophers have
tried to put these three components/conditions of knowledge into various forms and some
modifications have actually been provoked by the Gettier poser. The following samples of opinion
from different scholars reveal significant convergences despite differences in choice of words to
convey the same idea:
Clarence Irving Lewis (1883-1964) describes the following conditions for knowledge:
(a). Knowledge must be apprehension of a belief in what is true or a fact as against what is false or
is not a fact. An apprehension which is in error is probably called cognition but only cognition
which are true or correct should be classified as knowledge. Cognition as such is a process which
implicates the encounter (ignition) of both the faculty/capacity and the object of knowledge (hence
co-ignition) and it is not just the beginning of such a process but all the stages count as cognition
and knowledge is only a product of cognition.
(b). Cognition generally or (the content of it) must have meaning in the sense that something is
signified, believed in or asserted which lies beyond or outside the cognition experience itself. When
such cognition is veridical (i.e. true), or is knowledge, it must correspond, accord with or be true
of what is, thus (made or) affirmed.
(c). Knowledge must have ground or reason. (d) Knowledge (or at least knowledge in the best and
in the strict sense) must be certain of what is believed, if knowledge is no best than probable, then
the belief may be justified to some degree but the knowledge is not of the fact believed but the fact
that the belief is probable 2

John Hospers (1918-2011) distinguishes two senses of knowing: (a) weak sense and (b) strong sense.
In the weak sense, knowing implies the following conditions:
(i) one believes it.
(ii) one has good reason to base one’s belief.
(ii) one takes the belief to be true.
In the strong sense, knowing implies

1
Certain modifications to this classical or traditional definition of knowledge have intervened since Edmund Gettier’s 1963
article titled “Is justified true belief knowledge?” We will consider this Gettier poser later but let us for now concentrate
on these classical tripartite conditions for knowledge which has been widely accepted by the past main-stream philosophers
like Plato, Aristotle, Aquinas, Descartes, Locke and later Russell. Thus, JTB constitute what is called the traditional
analysis of propositional knowledge. Put in other words, the traditional tripartite analysis of knowledge focus on three
components, viz: belief, truth and justification.
2
cf. C. I. Lewis, An Analysis of Knowledge and Valuation, USA: Paquin Printers, 1946, pp. 315-365).

2
 (i) one believes it
(ii) one believes it to be true.
(iii) not only good reasons are demanded but also absolutely conclusive evidence.3

Alfred Jules Ayer (1910-1989), the father of British logical positivism holds that the role of philosophy is
to discover the meaning of things.He argues that there are necessary and sufficient conditions which
something has to fulfil in order to be called knowledge even though he did not believe that the
conditions can be fulfilled. Such necessary and sufficient conditions are;
(a) that what one is said to know be true.
(b) that one be sure of it.
(c) that one should have the right to be sure and this right can be earned in various ways such as
(i) valid proofs,
(ii) reliable authority,
(iii) reference to perception
(iv) reference to scientific laws
(v) reference to intuition
(vi) reference to historical records.
For Ayer, it will be difficult to give a complete description of knowledge because the right to be
sure (i.e. the logos condition) is divergent.4

Neal W. Klausner and Paul Grimley Kuntz hold that the basic minimum conditions for any claim to
knowledge include:
(i) acceptance, acknowledgment, and affirmation of a statement that something is the case.
(ii) that one has adequate, good and sufficient evidence that something is the case
(iii) that what is actually affirmed is the case.5

Peter Carruthers claims that for something to be called knowledge, the following conditions are necessary
(a) it must be believed to be true
(b) to the true belief is added a controversial quality, namely justification which is derived either
from a reliable source or some causation that obtains or that makes the belief to be imperative. 6

For Douglas Odegaard, ideally our knowing something involves four things;
(a) our being sure of it
(b) the certainty of what we are sure of; so knowledge involves a personal response; it makes
demands on our rational judgements
(c) Our being permanently justified in being sure of it. A knower must be justified, in claiming the
certainty of what he knows.
(d). The truth of propositions is essential to justification. Knowledge must require the truth of what
is to be known.

3
John Hopkins, An Introduction Philosophical Analysis (UK: Routledge, 1997), pp. 48-49.
4
A. J. Ayer, The Problem of Knowledge,(Bombay: Macmillan, 1957)
5
N. Klausner and P. Kunz, Philosophy: The Study of Alternative Beliefs (New York: Macmillan, 1961)
6
P. Caruthers, Human Knowledge and Human Nature: A New Introduction to An Ancient Debate, (Oxford: Oxford
University Press, 1992), pp. 2-3.

3
 Having had a panoramic account of a selected views on the conditions of knowledge, we have
to begin to consider each of the recurrent three conditions (justification, truth, and belief=JTB)
separately. Knowledge as we argue is the end product of the cognitive process. This process however
does not begin with justification rather it starts with assenting to some beliefs and then progress to
check if our beliefs are true (or false), and if true, then we ask about the reasons or justifications for
our true beliefs. Accordingly, we pose the first question: What then is the Belief (pistis) Condition for
Knowledge?

THE BELIEF CONDITION

The belief condition for knowledge can be represented as follows:
If ‘A’ knows that ‘P’, it is claimed then that ‘A’ believes that ‘P’ (KaP – BaP).
Basically, this formula KaP - BaP gives us a rough map of the relationship between believing and
knowing, without stipulating what constitutes this relationship.
Howsoever, we have to underscore, from Plato’s Allegory of the Divided Line (Republic 509d-
511e), that belief (Gk: pistis) is something lower than knowledge (Gk: episteme) but the question
which the belief condition tries to address is about the relationship between the two: has belief anything
to do with knowledge? For instance, what is the relation between the two statements “I believe that
God exists” and “I know that God exists”. This relation is what the belief theories or belief theses try
to explain.
Over the years, a number of theories have been developed which try to present this more
systematically. Of these, we can say that three theories or rather theses have been traditionally
proposed to explain the relationship between ‘knowledge and belief’: namely (i) the entailment thesis
(ii) the incompatibility thesis (iii) the separability thesis. Let us treat each of these theories separately:

(i) The Entailment thesis: According to many epistemologists, knowledge entails belief – i.e. I cannot
know that such and such is the case unless I believe that such and such is the case. For this
entailment thesis, if there is no element of belief in knowledge, then knowledge is a mere guess
work. Thus most epistemologists believe that belief is a necessary and logical condition for
knowledge, that is to say that knowledge is at least true belief. To avoid confusion, some
philosophers have demanded that belief be substituted with other related attributes, like conviction.
Those who thought of this substitution include A. J. Ayer, Neal Klausner and Paul Kuntz. Ayer
would prefer the phrase ‘being sure’; other substitutes may also include ‘certainty’ and ‘conviction’.
Of course, there are other epistemologists who argue that belief cannot be a condition for
knowledge. Such critics are the incompatibilitists.

(ii) The Incompatibility thesis: Plato would represent belief as a grade of cognition lower than
knowledge. For one cannot believe and know at the same time, he claimed. For we talk of belief
precisely when we do not know. To talk of belief when one knows is misleading because it
expresses only a part of what one wishes to express, namely he knows. One of the contemporary
defenders of this thesis is Harold Arthur Prichard (1871-1947). He says that knowledge involves
certainty either in the sense of infallibility or in the sense of psychological certitude. Thus, your
mind is firmly sure but according to him belief does not involve certainty7 So Prichard made a
distinction between knowledge as indubitable and belief as dubitable. Knowledge can be neither
true no false, only beliefs are. Therefore, believing something rules out the possibility of knowing

7
cf. Kleur, P. Certainty: A Refutation of Scepticism, Minneapolis: University of Minnesota Press, 1981).

4
 it at the same time8 Here we say that knowledge transcends belief. Belief means looking at
something vaguely and when knowledge sets in, a thing becomes clearer. Therefore, to make belief
an element of knowledge is to be redundant. Prichard’s objection can be answered by saying that
belief demands some level of confidence, that is to say, that Prichard is wrong by saying that belief
is a guess work. Secondly, belief might be a component of an infallible form of knowledge. For the
uncertainty and dubitability of belief will be compensated by other conditions. Another person in
this incompatibility thesis is the British philosopher Austin Duncan Jones (1908-1967) who said
that knowledge and belief are incompatible. He arrived at this distinction by the various linguistic
uses of ‘belief’ and ‘knowledge’ e.g. He said that people usually say - ‘I don’t believe (that) he did
it but I know he did it’ – i.e. to say ‘I know ‘P’, but I don’t believe ‘P’. One could recognize on
logical grounds that ‘P’ is true or acceptable but on emotional or psychological grounds, false; and
it is difficult (on pragmatic grounds) for us to accept it. There are also objections to this thesis. For
instance, some say that Jones is misrepresenting people’s statements – that I don’t believe he did it
but I know he did it’. What people are meaningfully saying is only an attempt to express a shock
and they are emphasizing that shock by saying – ‘I don’t believe it’. It is a matter of shock and not
an expression of incompatibility of knowledge and belief. Some people like Lehrer have said that
knowledge does not involve ‘certainty’ – in order to avoid scepticism, ‘acceptance’ could be better.

(iii) Separability thesis: This thesis holds that belief and knowledge are separable though they can
go together. This position is defended by A. D. Woozley (1912-2008) and Colin Radford (1935-
2001). Woozley said that knowledge can exist in the absence of confidence about the item known,
although knowledge can be accompanied by confidence (belief) as well. Even when people are
unsure of their claim, they might know that the claim is true. Though Woozley accepted that it
would be odd for those who lack confidence to claim knowledge. Radford, extending Woozley’s
claim, states that knowledge is not only compatible with the lack of certainty, it is also compatible
with the complete lack of belief e.g. John had forgotten that he studied the history of science in
philosophy one but he gave a correct answer to the question about when the scientific revolution
began when asked – this is a spontaneous answer though without believing that his answer was
correct. One can say that he knows it without believing that he knows it. Some people will object
to Radford by saying that John had a belief which is unconscious. It is the case in which John’s true
belief becomes unconscious but persisted long enough to cause the present correct guess answer.
Like Woozley, however Radford could also acknowledge that when we claim knowledge, we ought
at least, to believe that we have knowledge or else our behaviour is intentionally misleading.
In spite of their dissenting theses, most epistemologists hold that knowledge requires belief, but belief
does not mean knowledge. We ordinarily deny that someone can know that there are for example, nine
(9) planets without believing it. If, as many believe that, our knowledge arises in experience, in our
reflections, and inferences, then it is at least, believing with more factors added to it to upgrade it to
knowledge. Knowledge is a belief of a certain kind, satisfying certain conditions, which from the
traditional point of view, would include the truth of which he (the knower/believer) believed.
Philosophers have rarely disagreed that knowledge required belief but they have often disagreed on
the nature of belief itself. For this reason, some who deny belief to knowledge appear to do so based
on what they understand by belief.

8
cf. Prichard A. H. Knowledge and Perception, Oxford: Clarendon Press, 1950; see also G. E. Moore, Philosophical
Papers, N.Y.: Macmillan, 1959).

5
Let us now pose the second question: What then is the Truth (aletheia) Condition for Knowledge?
THE TRUTH CONDITION FOR KNOWLEDGE
The relationship between truth and knowledge is much more conspicuous than the relationship
between knowledge and belief, hence knowledge (episteme) as distinguished from belief (pistis) and
opinion/illusion (doxa/eikasia) or conjectures, must have a necessary ingredient of being true. To
know something, we must be correct in what we believe. Hence knowledge demands not mere belief
but a truth requirement.
If what one claims to know eventually turns out to be false, it simple means that one did not
know it. So, it is not the case then that astronomers before Copernicus know that the earth is flat. Since,
it is false that the earth is flat, no one knows that the earth is flat. Knowledge without truth is
impossible, according to the traditional conception of knowledge.
We are inclined to suppose that truth is the proper aim of scientific inquiry, that true belief
helps us to achieve our goal, that reliable preservation of truth (as one argues from premises to
conclusion) is the mark of valid reasoning. One condition for being sure that one knows something is
that what one claims to know must be the case, it must conform to what is, it must be true.
Philosophers generally agree that truth is required for knowledge, but they have offered some
opposing ideas about the question of what is it (it is) for a belief to be true. When is something or a
belief the case? When is a proposition true? How do we determine when truth is present? These
questions have always occupied epistemologists.
Plato for instance understands truth as a-letheia which means unveiling (or revelation) what is
already there in the ideal world but forgotten due to the soul’s union with the body. What the soul
forgets, it can however retrieve from memory hence his theory of anamnesis (remembering).
Heidegger, in the light of existentialism understands truth in terms of freedom (to search for
the essence of man), hence it is to be discovered or disclosed (a-letheia) in the course of man’s search
for authenticity, which by implication means freedom. Truth for him is thus understood in terms of
freedom.
Nonetheless, for epistemologists in general, there has been three traditional theories about the
nature of truth or what it is to be true. These three traditional theories include: 1. The correspondence
theory of truth 2. The coherence theory of truth 3. The pragmatic theory of truth. Of course, other
theories have been developed but these three remain the most traditional. At this stage of our study,
our concern is not yet with the detailed theories of truth, but with truth as a condition for knowledge.
However, since we cannot easily or fully understand these truth-conditions for knowledge,
without some preliminary understanding of truth, we must therefore begin with the consideration of
the nature of truth (even if schematic) in the light of these theories.

1. Correspondence theory: Since the time of Aristotle, many philosophers including Aquinas, Locke,
Moore, Russell, early Wittgenstein, G. L. Austin, and majority of philosophers and indeed traditional
philosophers have defended the correspondence theory. There have been a lot of versions of this theory
but the central thesis is that true proposition corresponds with reality or some may say, truth consists in
correspondence between sentences – like truth-bearers and features of the actual world. A candidate
for truth is true if and only if it corresponds to facts. Modestly put, some say that a proposition, sentence,
or statement is true if it represents reality – i.e. if at least it expresses a fact. Thomas Aquinas in the
classical tradition has put this as “veritas est adequatio rei intellectus” i.e. adequation of intellect to the
things. Nevertheless, the problem is to determine the exact sense in which a truth statement corresponds
to reality. What does this mean? Does it mean a literal picturing of reality as (picture theory)
Wittgenstein would say, or does it mean mirroring? Does it mean detailing word to word (copying)

6
 relationship with reality? Assuming that a correspondence theory of truth is a reality, how could it be
related to knowledge? How can one know that a correspondence relationship holds between a statement
and a world? Does such knowledge require that we compare a proposition with the way a world is, in
order to determine that the former corresponds with the latter? How can we compare something which
is in the realm of the mind with that which is in the realm of the world? How can one be justified in
believing that such a relationship holds? Do we have an independence of access to how the world is? –
a means that does not essentially involve acceptance of a proposition about the world? Some agree then
that if we already know that the world is (exists), it then means that we know the truth about it; and to
study it again will be redundant. In spite of all these problems, all the versions of the correspondence
theory will hold that a statement can be true because of its correspondence to the way things are (in
reality) even if no one knows that this relationship holds or obtains. And this is the normal generally
accepted position of epistemological truth. This theory is held more by the realists.

2. Coherence theory: Here a statement is said to be true if it co-hers with a specific (or comprehensive)
system of statements “vera verum consonant”. This is in the matter of ideas to ideas and propositions to
propositions. A statement is true if it fits in with other systems of truths or ideas. Coherence theory
consists in a relation which truth-bearers have with one another. This theory is seen among philosophers
who are system-builders. This is, in particular, the idealists’s position. People like Leibnitz, Hegel,
Descartes, Spinoza and other rationalists were interested in building systems of thought rather than a
system corresponding with reality. One has to make sure here that the ideas in-view co-hers with the
one already accepted. Even though coherence can be a mark of true theory, but (yet) they cannot be the
truth. This is because the thing to cohere with might be false. Here, to say that truth itself is coherence,
then it means that is problematic. Some say that coherence theory presupposes the correspondence
theory. Why must the recognition of such coherence be taken as adequate for knowledge of how things
actually are? Also people have agreed that there can be a system of consistent body of falsehood. This
means then that any attempt to seek for knowledge with reference to how things are, does not hold
water. If truth and knowledge are divorced from the way the world actually is, then knowledge will no
longer be what is. It also presupposes a kind of correspondence theory. This leads to an unaccepted
relation about truth since many different and mutually incompatible system of beliefs should be
internally consistent and self-supporting. This theory confuses the criteria for truth and the nature of
truth. Correspondence theory appears to be a better explanation of truth in this case for it is more
fundamental and presumed by other theories, if they are not to relapse into a dangerous relationism,
circularity, and infinite regress.

3. Pragmatic theory: This holds that a proposition is true if it is useful in a certain way. The proponents
include William James, John Dewey, Richard Rorty, Sanders Pierce. William James sees truth as
workability of ideas and Dewey holds that truth is success in the inquiry. Sanders Pierce sees truth as
conventionality. Dewey’s theory is the theory of instrumentality. This theory, especially in the hands of
William James, urges a connection between what is true and what is useful. Pointing out for instance,
that a mark of successful scientific theory is that it enables us, through associated development in
technology, to manipulate nature in ways hitherto unavoidable to us. The question then is: Is every type
of knowledge reducible to usefulness? There are a wide range of our knowledge that are true but not
useful. And when is something useful? Is another question. This is because what is useful in one case
may be useless in another sense. Therefore, to reduce truth to usefulness will not help knowledge. Critics
also point out that the confliction (conflation) of truth with utility is perilous because the ethics of belief
is required as to pursue the truth with honesty even if its consequences should prove detrimental to our
material well-being – i.e. we have to see truth in terms of knowledge, that is to say that we have to
understand (that) truth as a condition for knowledge in terms of truth relating to what is the case. For as

7
 Aristotle said, to say that “of what is, that it is and what is not, that it is not, is to say the truth. Truth is
a condition for knowledge but it must refer to what is.

4. Semantic theory: As the label ‘semantic’ implies, this theory of truth deals with meaning of sentences
or propositions by insisting that the satisfaction (iff) criterion of the truth-claims of propositions is basic
to truth-definition. It states that sentences have properties we call truth and so they are truth-bearers.
One could even say that this theory is a more developed version of the classical Aristotelian
correspondence theory of truth. Associated with this theory is the T-scheme (Tarski-scheme)
equivalence denoted as ‘if and only if’. This scheme is usually stated in logical expressions, e.g. A is
true, iff A. Hence a sentence of the form “A and B” is true, if and only if A is true and B is true; whereas
a sentence of the form “A or B” is true if and only if A is true or B is false. The T-schema is so called
because its proponent is Alfred Tarski. Tarski is a Polish philosopher and with his 1933 monograph
Pojęcie prawdy w językach nauk dedukcyjnych (The concept of truth in languages of deductive
sciences), he popularised the semantic theory of truth. Two years later in 1935 this book was translated
into German as Der Wahrheitsbegriff in den formalisierten Sprachen, and circulated amongst the
analytic philosophers of the Vienna Circle. The main thrust of this theory is with regard to truth as a
property of sentences as well as the relation between truth and meaning. Its use applies more to formal
languages than the ordinary language. Tarski tries to analyse and solve semantic paradoxes (e.g. the
Liar Paradox) as well as draw attention to the material and formal adequacy of a truth-definition by
highlighting the relation between a language and a metalanguage. Let us take German as the language
(L) and English as the metalanguage (ML): in semantic theory of truth, we can use English (ML) to give
the meaning (semantic value) of a German sentence – for instance the truth of the sentence “Schnee ist
Weiss” (L) can be given by as follows – “The sentence Schnee ist Weiss means ‘snow is white’” (ML).
The semantic theory of truth is mostly used in philosophy of language, logic and mathematics. It has
both critics (e.g. Ottoh Neurath, Hilary Putnam) and defenders (Rudolf Carnap, Karl Popper) amongst
analytic philosophers while its most recent reinterpreters or modifies include: Donald Davidson and
Saul Kripke. Gottlob Frege, one of the analytic philosophers who focus on the theory of meaning
(Bedeutung) makes a distinction in his semantic theory between ‘sense’ and ‘reference’. For him
“semantic theory is a theory about the determination as true or otherwise of sentences in accordance
with their composition out of words and phrases; and the semantic value of a constituent of a sentence
is that by which it contributes to determining the truth or otherwise of the whole sentence. That is why
(Frege takes it for granted that) anything that really contributes to determining the truth or falsity of a
sentence must have a Bedeutung (meaning), must have some semantic value.”9

5. The Deflationist theory: There are many variants of this theory which include the redundancy theory,
the disappearance theory, the minimalist theory, the no-truth theory and the disquotational theory. The
slight differences notwithstanding, we can say that in general proponents of deflationist theory of truth
argue against the semantic theorists by insisting that truth is not a property (shareable between two
propositions) and in fact they deny that truth has a nature. In other words, they desist from giving any
explicit definition of truth and in fact criticize the correspondence or coherent theories of truth. For
them, to assert that a statement is true is simply to assert that statement. For instant to say that ‘snow is
white’ is true or that it is true that ‘snow is white’ is simply equivalent to saying that snow is white. In
this way, deflationists are associated with the equivalent schema. They also make a distinction between
analytic (redundancy, disappearance, no-truth versions), material (disquotational version) and
necessary (minimalist version) equivalence.

9
Julian Baggini & Jeremy Stangroom (eds), What Philosophers Think, (London-New York: Continuum, 2003), p.218.

8
We now pose the third and most controversial question: What is the Justification (logos) Condition for
Knowledge?
THE JUSTIFICATION CONDITION FOR KNOWLEDGE
It has been observed that many true beliefs do not qualify as knowledge because true beliefs
can also qualify as guess work. It is not enough for a belief to be true to qualify as knowledge because
it is not impossible to have a lucky guess or a conjecture that happens to be true.
Socrates in Plato’s Meno, distinguished between right opinion and knowledge by a popular
metaphor called the statue of Daedelus. Daedelus, by the way was a distant relative of Socrates who
happens to be an artisan and a military engineer. He designed sea vessel for the Greek army (i.e Navy
Seals) and according to later mythology, he constructed wings of the vessel such that it can fly with
feathers attached to wax. He and his son were supposed to have flown from a high cliff but they flew
too near the sun, such that the wax melted and he fell into the sea. Moreover, Daedelus’s statues are
said to be life-life that they would run away if they were not chained down (cf. Meno 97b-98a;
Euthyphro 11c-e – 15b-c. Socrates claims that knowledge is like the statue of Daedelus, which if not
chained will also run away. Therefore, knowledge is to be chained down or anchored by justification.
A mere right opinion or (true) belief that falls short of knowledge is like an unchained statue
of Daedelus, fleeting, lacking in solid reasons, warrant. One can believe something that happen to be
true but without good reasons to support the belief. This thereby disqualifies it from being genuine
knowledge. So, for one to say that he knows, there must be some grounds or supporting reasons for
what he claims to know, i.e. justification, evidence, warrant.
Justification then specifies the nature of support which different kinds of belief must have if
their claims to knowledge must be made true. For it is possible to arrive at and be wholly convinced
by a belief that is in fact true on superstitious or absurd grounds (this is like the parable of the sower
about seeds that fell on sandy soil without depth. Consider fallen Catholics who run from one church
to another.) Hence some argue that knowledge is a true belief for which the true believer has sufficient
evidence or some such formular.
Justification by its nature has some kind of connection with the truth because it has a sort of
property which by its nature apparently is said to ground belief in the real world which we take to
count towards the truth of the belief and in ceteribus paribus (all things being equal) towards its being
knowledge.
Justification relates ‘belief’ and ‘truth’ to knowledge. For there to be knowledge it is necessary
that the satisfaction of the belief condition be appropriately related to the satisfaction of the truth
condition. What assures this appropriateness is justification. This mediating link between belief and
truth is justification which means offering reasons (or specifying what qualifies as reason) for our
beliefs. Epistemic justification is then a property ascribed to a belief in virtue of satisfying certain
evaluative norms concerning what a person ought to believe. Such norms measure the goodness of a
belief in so far as we are interested in epistemic goals – namely attaining truth and avoiding error10

10
cf. Ted Honderich (ed), The Oxford Companion to Philosophy, (New York: Oxford University Press, 1995, p. 434.

9
ROUTES TO EPISTEMIC JUSTIFICATION
There are broadly two routes of justification – internalism and externalism.
Internalism (e.g. deontologism, mentalism and access internalism) holds that the reason for
justification is within the cognitive sense of the individual, - within one’s introspective vision; whereas
externalism (reliabilism, causal theory of knowing,) holds that the reason for justification is a reality
outside (extra-mental) the knower.
The justificationist account of knowledge is traditionally one which subscribes to internalism.
Deontologism for instance implies an internalist self-supporting evidence for holding unto one’s
opinion as an epistemically true belief, hence it argues that one is justified in holding a belief as true
if and only if one is epistemologically responsible (deon = duty) for the finding the evidence as well
as has a direct voluntary control over his reasons for believing that p is the case at any given moment.
This is only one version of internalism but there are other versions. So, what does justification in the
internalist perspective demand?
Certainly, the classical view demands that justification involves a matter of adequacy of the
reasons or evidence one has for a belief. This traditional/classical view conceived justification as a
necessary condition for knowledge as well as sees it as a matter of adequacy of the reason or evidence
one has for a given belief. Traditionally, epistemic justification must satisfy certain sets of intuitions,
which has been identified as the following:
(a) justification is internal – It has to do with one’s cognitive set and not just with circumstances
independent of one’s cognitive set. Though internalism may not require that all of one’s reasons be at
the forefront of one’s mind or even that one be conscious of believing them in order for them to play
a justifying role.
(b) justification is truth-oriented. It is not just offering reason. If X is justified in believing that ‘P’,
then it has some reasons that ‘P’ is true as opposed to mere conviction, satisfaction and usefulness.
(c) justifying reasons are non-arbitrary. No belief can play the role of a justifier if it is merely
adventitiously contained in one’s cognitive sets. A justifier cannot be simply posited without the
concrete basis. “A belief acquired at random without evidential connection to the believer’s current
cognitive set has as such, no particular claim to truth…..A basic intuition regarding justification is that
no belief can pass on credibility unless it possesses some credibility itself. Hence, such arbitrary beliefs
cannot confer justification”11
This can be expressed in another way when it is said that “To have knowledge is to have a true belief
brought about by a non-accidental way. To say that a belief has been obtained in a non-accidental way
is to ensure that it is knowledge and not just a matter of possessing true belief. For a true belief to count
as knowledge, it cannot be an accident or a matter of luck that I happen to believe it. My having the
belief and the fact in the world must be connected in some way12

11
cf. Timothy, J. McGre, The Foundations of Knowledge, (Maryland: Little Field Adams Books, 1995), pp. 8-11.
12
cf. Michael Luntley, Truth and Self: The Post-Modern Reconsidered, (London: Routledge, 1995), p. 91.

10
PROBLEMATICS:
The traditional presentation of justification as a condition of knowledge presents us with some
problems, which will be a matter of full treatment later as we progress. Like the skeptics’ denial of the
possibility of justification, based on the problem of infinite regress and vicious circle namely:
(i) What justifies the justifier?
(ii) Is every justified belief based upon other beliefs that act as reasons or are there some
fundamentally justified beliefs that don’t demand other beliefs as justifiers (i.e.
coherentism or foundationalism)?
(iii) Are there foundational beliefs that are self-authenticating or self-warranting or self-
guaranteeing?
These are questions for further considerations, when we shall consider the skeptics. For now,
what preoccupies us is simply to find out whether the notion of knowledge demands justification and
if so, under what condition(s) are reasons for some beliefs good enough to make knowledge possible?
The question is provoked by the second (ii) and third (iii) questions above. Some have argued that
knowledge is merely true belief and no justification is either necessary or possible. Just like in truth, if
it is argued that justification of a belief that is true is redundant or inconsistent in that, justification
would have to be known to be true and so itself requires justification. Some say that ‘justification’ and
‘true’ are tautologous as they are arrived at through the same process.
All these questions, especially that of finding out the quality of reasons that confer knowledge
to belief will be the subject of the Gettier poser and the controversies emanating from it thereafter.

Justification, knowledge and the Gettier poser (a skeptic on justification):

We have already said that many philosophers assume that a justified-true-belief is sufficient as
well as necessary for knowledge. However, this definition has become very controversial due to what
has been called ‘The Gettier-poser’.
Edmund Gettier, as earlier noted, in 1963, published an influential challenge to the prevalent
view that if you have a justified-true-belief that P, then you know that P – by presenting certain counter
examples. Since after Gettier, other people have also elongated/widened/ provided more counter
examples namely that it is possible to have a justified-true-belief that does not constitute knowledge.
Let us offer some of these Gettier-inspired counter examples:
(1) A person inquiring from an office receptionist the time. The receptionist consults the office clock
which has been accurate for years and gives the time as indicated by the clock. The fact of the matter
is that the time read of is correct but unknown to the receptionist, the clock had stopped exactly twelve
hours earlier. Now, does it make sense to say that the receptionist knew the correct time at the time he
gave it out (suppose he discovers 10 minutes after the encounter that the clock had stopped even if he
had a justified-true-belief.) Can we say that the receptionist knew the time? Of course, ‘No’. because
his true belief is a pure coincidence.

(2) Smith and Jones have applied for the same job. It was the criteria for the job that the one who has
10 coins will get the job. Smith is justified in believing that (a) Jones will get the job (b) Jones had 10
coins in his pocket. On the basis of (a) and (b), Smith inferred that Jones is the person who will get
the job as he had 10 coins in his pocket. As it turned out, Smith himself actually got the job and it
happens also that he had 10 coins in his pocket. Although Smith is justified in believing the true
proposition that the person who will get the job had 10 coins in his pocket, yet Smith does not know it
(given that the basis/justification of his belief is the 10 coins in Jones’s pocket not the ones in his own
pocket).

11
 Gettier counter examples are cases where one has a justified true belief (a justified belief that
happens to be true/correct) and yet could not claim to have known. Therefore, an understanding of the
nature of knowledge must require a Gettier resistant account of knowledge. The Gettier problem is a
problem of finding the modification or an alternative to the traditional justified-true-belief analysis
that avoids the difficulties of Gettier-styled (Gettier resistant/Gettier-modified) counter examples. For
more than three decades of vigorous research, contemporary epistemologists have still not come out
with a widely-accepted solution to the Gettier problem.
Attempted solutions have revolved around two axes:
(1) Attempts to add another element or condition to the justification agenda – something that will help
to resist the Gettier-styled counter examples.
(2) To bring another solution beyond the justification agenda, i.e. justification, truth, and belief
condition for knowledge. There can be a third axis which holds that “Gettier-counter examples are
faulty because they are calling what is not knowledge, knowledge (those people in Gettier counter
examples have no knowledge initially, i.e. in the first instance).
As the discussion of the responses and/or solutions to the Gettier-poser demands our full attention, let
us consider the two routes to its solution (internalist and externalist), even if not exhaustive:

The INTERNALIST Response to the Gettier Problem: Internal here refers to mental attitude or
subjective states like ‘belief’. Our beliefs are simply internal (mental) attitudes and if true, they
approximate to knowledge, hence the internalist response emphasizes the quality of the reasons that
distinguishes knowledge from true belief. In other words, it tries to consolidate the justificatory
conditions for our true beliefs by adding some other condition (a fourth condition, so to speak – in
addition to JTB), so as to serve as a buffer against possible ‘Gettiered counter-examples’ – We have
three such internalist justification responses, namely:

(1) The Defeasibility condition: This was even suggested by Gettier himself and also by many others.
For this group, to justification must be added, “the indefeasibility condition” (i.e. defeasibility
argument). For something to be a condition, the justification must be undefeated. i.e. whatever you use
for justification should in no way be defeated. For something to be a condition, it can’t be defeated.
Justification appropriated to knowledge must be undefeated. A justification is defeated when it is based
on falsehood, false presuppositions or if the evidence the individual possesses is counteracted by other
evidence that may come up. We require, for knowledge, that justification be indefeasible. Defeasibility
could further be characterized in terms of evidence that the subject does not possess (imara nke a, i ma
nke ozo?) which overrides or defeats the subject’s prima facie justification for belief. People say if
defeasibility is added, (i.e. undefeated justified true belief = UJTB) there are things that are always put
to some doubt. It is a very difficult condition for knowledge.

(2) Non-defective condition: Some people say we have to add non-defective element. That
justification should not be defective (non-defective, justified true belief = NJTB). Roderick Chisolm
distinguished between defective and non-defective types of knowledge. Defective justification occurs
when the evidence that justifies a subject in accepting a certain belief also would justify the subject in
accepting at least one false belief, i.e. the evidence you use in justifying a belief has potentiality of
misleading. Non-defective justification contrastly occurs when the evidence doesn’t imply or entail or
lay epistemic support to any false belief, i.e. if there is a false belief, it isn’t because of evidence that
gives support to the false belief. The traditional definition of knowledge in this perspective will be a

12
non-defectively justified true belief.13This condition imposes a stronger requirement to distinguish
truth from falsehood.

(3) Conclusive Condition: These people say that knowledge is conclusively justified true belief
(=CJTB). This means belief is justified in that whatever justifies it guarantees truth. i.e. whatever
means you use to justify it is automatically true. Knowledge according to this opinion is associated
with certainty. Here we mean a certainty which a proposition has when there are extremely strong
grounds for it. Grounds that guarantee the truth. The problem here is trying to interpret what is
understood by certainty because certainty could be viewed in many ways. Does it mean indubitability,
which is complete subjective assurance where one is so certain that there is no way he can doubt it?
Does it mean infallibility, i.e. immune from error or does it mean irreversibility i.e. immunity from
repeat/revision? Does it mean self-authenticating or warranting? Does it mean incorrigibility? Does
it mean logical necessity? (People like Wittgenstein, Moore, and their likes will call it logical
necessity). If you make it indubitability, incorrigibility, infallibility and irreversibility, then most of
the what we call knowledge will not be knowledge, e.g. argument of authority or human weaknesses.
Today, people add things like appropriately-justified-true-belief (=AJTB) or the belief justified
rightly.

The EXTERNALIST Response to the Gettier Problem: Unlike the internalist response which tries
to anchor justification grounds on mental states (associated with ‘belief’) and the distinction of
knowledge from true belief, the externalist attempts to circumvent the whole question of justification
completely. By so doing, this solution aims at removing the very cog in the wheel at which ‘Gettiered’
counter-examples rest – in other words, it settles the matter by literally ‘running away’ from Gettier.
(To know, I do not need to believe nor do I have to worry if my belief is true or false, and so on –
justification is a non-issue).
This position could not require us to justify knowledge. Here knowledge is seen as simply
registering the truth just as thermometer registers temperature. Knowledge is analysed as in the natural
sciences where things are not understood by appealing to value-laden notions like justification, but in
terms of physical, chemical and biological and psychological properties together with the causal
relationships among them. These as we have seen are opposed to internalist approach where
justification is required for something to be knowledge.
According to Alvin Goldman, the externalist looks for justification outside mental or
phenomenological but typically in causal connection that obtains between the world and the subject’s
sense organs and central and peripheral nervous systems.14 The naturalistic approach is a general
umbrella that encompasses some related externalist theories like reliability theories (reliabilism),
causal theories and social coherent theories. We shall discuss the first two: reliabilism and causal
theories:

(1) Reliabilism: This theory diverts our attention away from the relationship between the propositions
claimed as knowledge and other false beliefs which should have been true or other truths which should
have been believed. In this reliability theory, it is suggested that a true belief can be knowledge when
it is derived from a reliable method. Early proponents of reliability theory include John Watling (1964),
Peter Unger (1968), D. M. Armstrong (1973) “truth-maker”; Alvin Goldman (1973) – its great
proponent. Reliabilism about knowledge would hold that a true belief is knowledge if and only if it is
produced by a reliable psychological process. That is, a process that produces a high proportion of true

13
cf. Roderick Chisolm, Theory of Knowledge, 3rd ed., (Englewood Cliff, New Jersey: Prentice Hall, 1987), pp. 90-100.
14
cf. Alvin Goldman, “A Causal Theory of Knowing” in Journal of Philosophy, 1967, p. 64.

13
beliefs. A process that typically yield truth. Thus, they are concerned with the reliability of the process
and not what the process offers at the end as knowledge. The demand here is that the means by which
the true belief is produced or maintained should be reliable.15 Drawing, for instance, from science,
some reliabilists would say that knowledge appears to be concerned with beliefs that are justified
because they are established by scientific inquiry. Even though they may be superseded by latter
developments in science, Goldman suggested that a true belief can be knowledge when derived from
a reliable method.16 Reliabilism’s appeal stems from the fact that it reminds us of the functioning of
epistemic agent without throwing overboard the criterion of knowledge. The various theories of
reliabilism seem to have in common reliable cognitive process. If you regard perception as a reliable
process, then once you perceive, you have a knowledge as long as your senses are sound. The senses
could include sense perception, memory, reasoning, empirical/scientific processes depending on
whether it has been considered a reliable cognitive process. The task of naturalized epistemology is to
characterise the condition in which a reliable process could be got. Commenting on the common
demands of a reliable cognitive process, Heil says “according to the theory, X’s belief would be
warranted, however he arrives at it, if it should happen that there is a sufficiently long period in the
future, whether or not we live to know about it, when belief generating process like X’s do (would)
yield true beliefs with a suitably high frequency”. Van Cleve citing or paraphrasing Goldman identifies
the common elements for reliable cognitive processes as follows: we say that a belief is justified if it
either (i) results from a reliable source (ii) results from a justified belief, via a reliable process.17

Objections to reliabilism: Reliability theory like other externalist theories is fraught with difficulties.
It has been pointed out that reliability theories are vague and typically failed to spell in any interesting
or persuasive way. Is it not possible that unreliable process may be accidentally lucky until now and a
reliable process unlucky until now? How reliable must a process be before it grounds knowledge? In
short, what is the measure of reliability? Equally many internal and external factors may affect
reliability making it to vary, e.g. inattention of the perceiver, distance of the object, the description
problem. How can we ensure an adequate description of reliable belief-producing process? How can
we choose between alternative description of the process.

(2) Causal Response: Here it is believed that when a true belief has an appropriate causal history, then
the belief counts as knowledge. A causal theory of knowledge is a form of externalism, and it is based
on the fundamental idea that a person knows some proposition P, only if there is an appropriate causal
connection between the state of affairs that makes P true and the person’s belief in P. The role of
causation in producing our knowledge is in the case of perceptual beliefs caused by the perceived
objects. Here, whether true belief is absolutely justified as an item of knowledge is exclusively the
function of whether there are suitable causal connection between one’s belief and external state of
affairs. BaP is a case of KaP, if P is true and furthermore the situation that makes P true is causally
responsible for the existence of the belief state BaP. My belief that P is true is a case of knowledge if
the conditions that are causally responsible for making the belief in me is true. For example, I am in a
room; I believe that I know that the room is hot because of the excessive heat in the room that caused
me to have the belief. It is no longer a mere belief but rather a caused belief. The epistemologists have
always recognised the importance of causal process in accounting for our knowledge of things e.g.
perception, memory, and reasoning are fundamentally causal means of knowing. Peter Caruthers
claims that for something to be called knowledge, it must be believed to be true and to this belief is
added a controversial quality – justification. And this justification is derived either from a reliable

15
cf. Peter Klein, “Epistemology” Concise Routledge Encyclopedia of Philosophy, (London, NY: Routledge, 2000), p.
248.
16
cf. A. Goldman, Journal of Philosophy, vol. 173, 1996, pp. 771-799.
17
cf. Jane Durrant, “Reliablism, Foundationalism, and Naturalized Epistemic Justification Theory” in Perspectives in
Philosophy, ed. Michael Boylam, Forthworth Harcourt Press, Javanowich Pub., 1993, p. 254.

14
process or some causation that obtains and which makes the belief to be imperative. On this causal
theory (or doctrine), knowledge is suitably caused belief. So belief is knowledge because it is caused
in a way that guarantees the truth.

Objections/Criticisms to causal theory: This causal theory can be criticised thus: (a) Not every case
of knowledge involves the situation where the thing or situation known is causally responsible for the
existence of the belief hence the causal theory is too restrictive. The causal theory cannot account for
a prior knowledge nor can it account for the futurabilia. (b) What if the person who has a natural phobia
for things not because the thing is fearful but because the person is phobic himself, thus it will lead to
more probability not knowledge because certain personal factors affect different people. (c) Causal
theory will make knowledge mere causal explanation of knowledge and so a non-serious activity rather
than a behavioural activity. The aspect of intentionality necessary for knowledge will be lost thus man
becomes almost a mere machine.

Conclusions
What we have seen so far from all the foregoing discussions on the normative approach to
knowledge is that knowledge is a complicated reality, very fundamental and very difficult to determine
the conditions for its possibility.
This difficulty can already be seen on the various conditions of knowledge listed out by Alfred
J. Ayer. According to him, in knowing, what one knows must be true and that one must be sure of it
(subjective certitude) as well as have the right to be sure. And this right to be sure can be determined
by a number of factors like (a) valid proofs (b) reliable authority (c) reference to percepts (d) reference
to reason (e) reference to intuition (f) reference to historical records. These numerous conditions made
him to pessimistically conclude that it will be difficult to satisfy the conditions for knowledge.
We do not intend to be as pessimistic as Ayer or to be brow-beaten by the difficulties and at
the same time quite recognisant of the problems. Thus, we think that the traditional approach to
knowledge as justified true belief, of course with some necessary modifications on its justification
condition, remains valid. And this remains the best way forward if we must beat the skepticist (e.g.
Pyrrho) and the naturalist (e.g. Quine) at their own game. This is why it may be said that knowledge
could be seen as appropriately justified true belief or even as some say that knowledge is true belief
justified in the right way or manner.
Equally, it could be said that the normative approach could be reconciled or integrated with the
other approaches, e.g. the descriptive approach to arrive at a more robust account of the conditions for
knowledge. To this end, we could propose such a long-winding ‘definition’ of knowledge as a
conscious true belief about a state of affairs engendered by facts, events, processes, information, or
ideas which ordinarily would be a right, correct and true reflection (expression) of the actual state of
affairs to be known, and of which the contrary would be a remote possibility if not an extraordinary
course of events.

Texts for Further reading on Knowledge as Justified True Belief

Classics:
Plato, Meno, Thaetetus.
Aristotle, On Interpretations
Aquinas, Summa Theologiae, I, qq. 1; 16-17; 84-86; II-II, qq.1-7.
J. H. Newman, An Essay in Aid of a Grammar of Assent (chapter 6)

15
Books:
Chisolm, Roderick, Theory of Knowledge, 1996, p.23.
M. van der Schaar, ed., Judgement and the Epistemic Foundation of Logic, Springer, London 2013
McGrew, Timothy, McGrew, Lydia, “Chapter 1: Internalism and the Collapse of the Gettier-Problem” in
Internalism and Epistemology: The Architecture of Reason, London: Routledge, 2007, pp.7-34.
Nagel, Jennifer, Knowledge: A Very Short Introduction, Oxford: Oxford University Press, 2014, esp.p. 58.
Nozick, Robert, Philosophical Explanations, Harvard University Press, 1981.
Peirce, C. S. “Truth and Falsity and Error” in J. M. Baldwin (ed) Dictionary of Philosophy and Psychology,
vol. 2, 1901, pp. 718-729.
Plantiga, Alvin, Warrant: The Current Debate, Oxford: Oxford University Press, 1992.
Pollock, John, L, Cruz, Joseph, Contemporary Theories of Knowledge, 2nd edition, Rowman & Littlefield
Publishers, 1999.
Popiel, Maximilian, A Systematic Review of Studies Using Gettier-Type Thought Experiments, (Thesis).
University of Denver
Scheffler, Israel, Conditions of Knowledge: An Introduction to Epistemology and Education, Chicago:
Scott, Foresman, 1965.

Journals
Alai, Mario, “Subjective and Objective Justifications in the Solution of Gettier’s Problem” in Logic and
Philosophy of Science IX (1), 2011, pp. 493-501.
Alvin Goldman, “A Causal Theory of Knowledge” in The Journal of Philosophy, vol. 64, 1967, pp. 357-
372.
Dretske, Fred, “Conclusive Reasons” in Australasian Journal of Philosophy 49 (1), 1971, pp.1-22.
Edmund, Gettier, “Is Justified True Belief Knowledge?” Analysis, 23 (6), June 1963, pp.121-123.
Lehrer Keith, Paxson Thomas Jr, “Knowledge: Undefeated Justified True Belief” in The Journal of
Philosophy 66 (8), April 1969, pp. 225-237.
Levin, Michael “Gettier Cases without False Lemmas?” in Erkenntnis 64 (3), 2006, pp. 381-392.
Richard Kirkham, “Does the Gettier Problem Rest on a Mistake?” in Mind, 93, 1984.
Skyrms, Brian, “The Explication of ‘X knows that p’” in The Journal of Philosophy, 64 (12), June 1967,
pp. 373-389.
Swain, Marshall, “Epistemic Defeasibility” in American Philosophical Quarterly. II (1), University of
Illinois Press, Jan. 1974, pp. 15-25.
Weinberg, Jonathan M., Nichols, Shaun, Stick Stephen, “Normativity and Epistemic Intuitions” in
Philosophical Topics 29 (1-2), Fall 2001, pp. 429-460.
Zagzebski, Linda, “The Inescapability of Gettier Problems” in The Philosophical Quarterly, 44 (174),
1994, pp. 65-73.

16