You are on page 1of 10

Argumentation and Interpersonal Justification

ALVIN I. GOLDMAN

University of Arizona
Department of Philosophy
Soc. Sci 213
Tuscon, AZ 85721
USA

ABSTRACT: There are distinct but legitimate notions of both personal justification and inter-
personal justification. Interpersonal justification is definable in terms of personal justifica-
tion. A connection is established between good argumentation and interpersonal justification.

KEY WORDS: Argument, argumentation, epistemology, justification, personal factors, social


factors

This paper concerns the relationship between argumentation, construed as


a social or interpersonal process, and the epistemic notion of justification.
According to the main tradition of epistemology, the requirements for
having a justified (or warranted, or reasonable) belief are individual rather
than social factors. One’s own perceptual experiences and prior beliefs,
for example, are the critical factors that make one justified or unjustified
in believing propositions like ‘A cat is present,’ or ‘Candidate X will win
the upcoming election.’ Such experiences and beliefs are events in the mind
of the believer, and are therefore categorized as individual rather than social
factors. In more recent times, however, some philosophers have proclaimed
a fundamental and pervasive role for the social dimension in justification.
Rorty, for example, contends that ‘justification is not a matter of a special
relation between ideas (or words) and objects, but of conversation, of
social practice’ (Rorty, 1979: 170). This suggests that social relationships
– presumably including argumentative relationships – lie at the heart of
epistemic justification. Since it is generally agreed that justification is a
necessary condition for knowledge, knowledge too would be invested, on
this view, with a thoroughly social character. This leads to what I shall
call Thesis 1:
(1) All justification, and hence all knowledge, is a fundamentally
social affair.
Although Thesis 1 is a radical view in some respects, it is less radical than
a second position often associated with a social approach to epistemology.
This second position accepts Thesis 1 in agreeing that if there were such
a thing as justification, it would have to be a social affair. But the position

Argumentation 11: 155–164, 1997.


 1997 Kluwer Academic Publishers. Printed in the Netherlands.
156 ALVIN I. GOLDMAN

I have in mind holds that no social process, even argumentation or per-


suasion, can produce genuine justification. All social-rhetorical processes
are purely ‘political’, in a sense that undermines or obliterates any claim
to rational justification. Let us call this Thesis 2:
(2) No social discursive process can create or transmit justification,
because all such processes (including argumentative processes)
are purely political and/or negotiatory.
I wish to deny both Theses 1 and 2. Against the first thesis, I contend that
there is such a thing as personal justification (P-justification), which is an
individual matter, and that a cognitive agent can be personally justified
(P-justified) in believing a proposition without having any relevant justi-
ficational relation to other people. Against the second thesis I maintain that,
in addition to personal justification, there is also interpersonal justifica-
tion (IP-justification), in which one person helps make a proposition be
justifiably believable by a second. Not all modes of persuasion are modes
of IP-justification, but some are. Rejection of Theses 1 and 2 is, I think, a
very commonsensical and unsurprising position. The larger and more
constructive task is to trace the connections between P- and IP-justifica-
tion in a satisfactory way, and more specifically, to explain how these
concepts relate to appropriate notions of ‘good argument’ and rules for
argumentation.
Although I maintain that there is such a thing as personal, nonsocial
justification, I cheerfully concede that the concept of epistemic justifica-
tion, like most concepts expressed in language, is socially derived or con-
structed, in some suitable sense of these terms. In claiming that there is
such a thing as P-justification, I am not talking about the source or origin
of this concept, but rather its content, that is, its conditions of satisfaction
or fulfillment. To take an analogy, we might grant that the concept of
baldness is socially derived and still insist that the conditions for baldness
pertain exclusively to the state of the individual, independent of his relation
to community or peers. Similarly, in saying (contrary to Thesis 1) that there
is a core nonsocial notion of justification, I mean that a belief can be
justified in virtue of such conditions as the believer’s perceptual experi-
ences and/or prior beliefs, rather than his/her relations to other thinkers.
Although there is no time here for a full-scale investigation of this issue,
I shall illustrate the type of social view that fits Thesis 1, and explain why
it is unsatisfactory.
One view that fits Thesis 1 is (1a):
(1a) Person S is justified in believing proposition Y if and only if
there are arguments S can present to his/her peers that would
persuade them of Y.
Under (1a), justification would obviously be a social affair, for it would
depend upon a person’s ability to enter into a relation of persuasion with
ARGUMENTATION AND INTERPERSONAL JUSTIFICATION 157

others in their community. An immediate problem with (1a) is its ambi-


guity. What proportion of one’s peers must one be able to persuade in order
to have justification? Does justification depend on having arguments that
would persuade all your peers? Most of your peers? At least one of your
peers? Clearly, this ambiguity would have to be resolved before (1a) could
be accepted.
Even without resolving this ambiguity, however, we can see that (1a) is
in trouble. Its first problem is that it does not provide a sufficient condi-
tion for justification, even in its strongest form. The strongest form of (1a)
says that S is justified in believing Y if and only if there are arguments S
can present to his/her peers that would persuade all of them of Y. A bit of
reflection, however, quickly reveals that satisfying this condition does not
entail that S is justified in believing Y. First, suppose S herself believes Y
as a result of accepting an argument in which Y is the conclusion and W
and X are the premises. Suppose further that the relation between these
premises and this conclusion is neither a valid one nor a strong inductive
one; in fact, the inference from W and X to Y is a flagrant fallacy. Then,
assuming also that this argument constitutes S’s only reason for believing
Y, S is not justified in believing it. Nonetheless, we can readily imagine
that S is capable of persuading all of her peers of Y, by use of this very
same argument. Suppose S is a very charismatic speaker, so whenever she
speaks, everyone is persuaded by all the arguments she offers. Then there
is an argument S could present to her peers that would persuade all of them
of Y; and yet, ex hypothesi, S is not justified in believing Y.
A different kind of counterexample to (1a) is one in which the infer-
ence from the premises to Y is a perfectly valid inference but S is unjus-
tified in believing the premises. Here again S would be unjustified in
believing Y (at least if these premises are her sole grounds for believing
it). But we may still suppose that if S offered the same argument to her
peers, all of them would be persuaded of Y. That still would not make S
herself justified in believing it.
Still another type of counterexample is one in which S could persuade
her peers of Y by means of a valid argument containing W and X as
premises, but S does not herself believe the premises, although she could
present the argument to others as if she believed them. Her own rationale
for believing Y is entirely different, and involves extremely poor reasoning.
Then S is not justified in believing Y, although she satisfies even the
strongest form of (1a) with respect to Y.
In addition to showing that (1a) does not provide a sufficient condition
for justification, our discussion provides ample reason for suspecting that
the core notion of justification is personal rather than social or interper-
sonal. Being justified in believing Y does not depend on how you would
fare in persuading others of Y, but on whether, for example, you yourself
believe premises that have an appropriately strong support relation (deduc-
tive or inductive) to Y. This is an essentially nonsocial condition.
158 ALVIN I. GOLDMAN

My claim that the core notion of justification is personal needs clarifi-


cation. P-justification, as I understand it, does not exclude cases in which
a person’s justification for a belief derives from the assertions of other
people. For example, your justification for believing the proposition
‘Candidate X will win the upcoming election’ might derive from reading
a newspaper report of an election poll. This can still be subsumed under
the category of P-justification because it is your belief that such a report
was in the newspaper that makes you justified. In other words, the justifi-
cation-conferring factor is an individual factor because it is a belief of the
cognitive agent. True, in this case the content of the premise belief concerns
other people or social agencies. But this is still just a special case of the
justification-conferring factor being a state of the cognitive agent.
If this kind of case can be subsumed under P-justification, one might
wonder whether there is really any difference between P-justification and
IP-justification. Indeed there is. IP-justification requires that all justifica-
tion has its source in the social or interpersonal arena, whereas P-justifi-
cation makes no such requirement. Thus, the concept of P-justification
would admit the possibility that young children before the onset of language
should have justified perceptual beliefs (e.g., ‘There is a toy under the
table,’ expressed in the child’s language of thought, not in a public language
which by hypothesis the child has not yet learned). Thus, P-justification
would allow for many cases of justified belief which IP-justification would
have to exclude. There would presumably be other differences as well
between P-justification and IP-justification, but this difference should
suffice for present purposes.
As the previous paragraph implies, IP-justification theories will tend to
fail by placing overly restrictive conditions on justified belief. In other
words, they impose conditions that are not really necessary for justifica-
tion. Let us see how this happens in the case of (1a). Consider the weakest
version in which (1a) says that a person is justified in believing Y if and
only if there is at least one (other) person she can persuade of Y. Even this
is too strong. S might not be able to mount a persuasive (and cogent)
argument for Y to anyone, because her own premises for justifiably
believing Y are ones she cannot get anyone else to accept, perhaps because
everyone wrongly and unfairly thinks that S is a liar. This unfounded
reputation as a liar renders S incapable of being persuasive to her peers,
but it does not render her belief in Y unjustified (cf. Goldman, 1994: 39).
Of course, the failure of (1a) does not prove that no correct social
account of justification can be given. After all, quite different sorts of social
theories might be proposed and considered. But our discussion clearly
displays the kinds of pitfalls that a fundamentally social approach encoun-
ters. It shows why I am dubious about Thesis 1. What I want to suggest
next is that the opposite approach, the approach that takes P-justification
as fundamental, is far more promising. If we take the fundamental or core
ARGUMENTATION AND INTERPERSONAL JUSTIFICATION 159

notion of justification to be personal, we can then easily explain one or


more notions of IP-justification in terms of the core, personal concept.
My procedure will be as follows. First, taking the notion of P-justifica-
tion for granted, I shall define a certain conception of a ‘good argument’.
Second, I shall use that notion of a good argument to define the concept
(actually several concepts) of IP-justification.
There are several viable conceptions of a good argument. One such
notion, the ‘logical’ notion, might be defined as follows:
(3) An argument is a good argument if and only if (i) the relation
between its premises and conclusion is either that of deductive
validity or strong inductive support, and (ii) it contains only true
premises.
While this is one legitimate conception of a good argument, it is not
satisfactory for all purposes. Richard Feldman objects to this notion with
the help of the following considerations (cf. Feldman, 1994). Suppose I
have been on a canoe trip in uninhabited wilderness and have been com-
pletely out of contact with society. I know that there was an election while
I was away and that my neighbor voted. However, I have no information
about whether she voted for candidate A or candidate B. Now consider
these two arguments:
Argument 1
1. Either my neighbor voted for A or she voted for B.
2. She did not vote for A.
3. Therefore, she voted for B.
Argument 2
1. Either my neighbor voted for A or she voted for B.
2. She did not vote for B.
3. Therefore, she voted for A.

Both of these arguments are valid, and one of them has true premises. So
one of them satisfies the conditions of (3) for being a good argument. But
neither is a good argument, says Feldman, since I have no reason to accept
either conclusion. Or, even if being valid gives them some merit, neither
is any better than the other, even though one in fact has true premises.
Hence the logical conditions for being a good argument are not sufficient
for argument goodness. Feldman goes on to claim that crucial to the
goodness of an argument is the epistemological or epistemic status of the
premises for the argument user. Although I do not fully agree with Feldman
in faulting the logical notion of good argument, I do agree that an episte-
mological sense of good argument is (also) extremely important, and I shall
follow Feldman in making use of his epistemological account.
Feldman offers roughly the following definition of the epistemological
conception of a good argument:1
160 ALVIN I. GOLDMAN

(4) An argument is a good argument relative to person S if and only


if:
(i) S is justified in believing the conjunction of all the
premises of the argument,
(ii) the argument is either valid or inductively strong, and
(iii) S is justified in believing that the premises are ‘properly
connected’ to the conclusion.

On this definition, neither of the two arguments about my neighbor’s vote


is a good argument for me (in the circumstances of my having just returned
from the wilderness). This is because in neither case am I justified in
believing the conjunction of all the premises. What is noteworthy about
Feldman’s epistemological conception of a good argument is that it is
relativized to a person and a time. Thus, an argument might be good relative
to one person but not relative to another, if the first person is justified in
believing its premises while the second is not. And a given argument might
change its status relative to a person when he gains more information
relevant to its premises.
By relativizing goodness of arguments to persons, definition (4) natu-
rally invites a personal interpretation of the justification-condition in
(4). If we think of a person’s justification as a matter of the evidence
possessed by the person, and if evidence-possession consists of the person’s
having certain beliefs and/or perceptual experience, then we have a personal
interpretation of the justification condition, an interpretation that readily
comports with the assumption that one person might be justified in
believing a given argument’s premises while another person might not be
so justified. In what follows, I shall understand definition (4) to presup-
pose such a personal conception.
Given the notion of P-justification, and the (epistemological) definition
of a good argument (which itself appeals to P-justification), how can we
define IP-justification? The most straightforward case in which a speaker,
S, justifies a proposition Y to a hearer, S*, is a case in which S presents
an argument to S* of which Y is the conclusion. Furthermore, in order
that proposition Y should be justified to S* by means of such an argument,
the presented argument must be a good argument relative to S* (in the epis-
temological sense of ‘good argument’). Among other things, S* must be
(P-)justified in believing the conjunction of its premises, and in believing
the premises to be properly connected to the conclusion, Y. S*’s justifica-
tion for believing the premises could stem either from independent sources,
distinct from S’s assertion of the premises, or from S’s trustworthy asser-
tion of them.
I think that the foregoing conditions capture the heart of IP-justification.
There are, however, a variety of slightly different senses or forms of IP-
justification that are worth distinguishing. I shall consider some of the
variations in abbreviated terms. First, is it enough (for IP-justification) that
ARGUMENTATION AND INTERPERSONAL JUSTIFICATION 161

S* be justified in believing the indicated things (the conjunction of the


premises and the proper-connection relation), or must she actually believe
them? Similarly, must S* actually believe Y as a result? I think that the
phrase ‘S justifies Y to S*’ is ambiguous (or vague) in these respects. There
is a persuasion-entailing sense of the phrase, which requires that S* believe
the conclusion as a result of believing the premises and appreciating the
premises-conclusion relation, and a second sense that does not entail that
S* be persuaded. Let us now formulate one of these senses of IP-justifi-
cation, the persuasion-entailing sense:

(5) Speaker S IP-justifies proposition Y to hearer S* (in the per-


suasion-entailing sense) if and only if:
(i) S presents an argument A to S*, of which Y is the con-
clusion,
(ii) argument A is an (epistemologically) good argument
relative to S*, and
(iii) S* comes to believe Y by inference from the premises of
A and appreciation of the proper connection between
premises and conclusion.

There is also a much weaker sense of ‘S justifies Y to S*’, viz., a sense


in which S* thinks that the argument given by S is epistemologically good
relative to her (S*), but in fact it isn’t. This is what we might call the
subjective sense of IP-justification. But I shall continue to focus on the
sense we have already isolated, which I would call the objective sense of
IP-justification.
Some further variations on the central theme concern the speaker’s status
vis-a-vis the argument presented to S*. Thus far we have required that the
argument must be (epistemologically) good relative to S*, the hearer. But
perhaps it must also be good relative to S, the speaker. Perhaps S too must
be (P-)justified in believing the conjunction of the premises, and justified
in believing that the premises are properly connected to the conclusion.
Furthermore, we should ask whether the speaker must actually believe these
things. If none of these requirements is imposed, then the speaker might
IP-justify Y to the hearer without the speaker herself being P-justified in
believing Y. Is this permissible? Again, we have several possible senses
of ‘S IP-justifies Y to S*’. In particular, there is a justification-creation
sense of IP-justification, which only requires that the speaker create (P-
)justification in the hearer, and a justification-transmission sense of IP-
justification, which requires that the speaker herself have (P-)justification
for Y, which she transmits to the hearer. Definition (5) expresses the jus-
tification-creation sense of IP-justification (combined with the persuasion
interpretation). The justification-transmission sense of IP-justification can
now be formulated by a slight expansion of (5) (in clause (ii)) to produce
the following definition:
162 ALVIN I. GOLDMAN

(6) Speaker S IP-justifies proposition Y to hearer S* (in the justi-


fication-transmission sense) if and only if:
(i) S presents an argument A to S*, of which Y is the con-
clusion,
(ii) argument A is an (epistemologically) good argument
relative to both S and S*, and
(iii) S* comes to believe Y by inference from the premises of
A and appreciation of the proper connection between
premises and conclusion.

We are now in a position to address Thesis 2. This thesis says that no social
discursive process can create or transmit justification. But definitions (5)
and (6) that we have constructed clearly undercut this thesis. They provide
clear senses in which a speaker might justify a proposition to a hearer. As
long as it is possible for an argument to be (epistemologically) good relative
to a person, then justification-creation and justification-transmission are
possible social relationships. As long as there is such a thing as P-justifi-
cation, and such things as ‘proper connections’ between premises and
conclusions (whether deductive or inductive), the indicated relationships
can obtain. Admittedly, the existence of such things is not wholly uncon-
troversial. But if these things are granted, the interpersonal conception of
justification (or several such conceptions) is readily generated.
IP-justification is not equivalent, of course, to mere persuasion (even in
the case of the persuasion-entailing sense of IP-justification). A speaker
might persuade a hearer of a proposition with an argument that is not epis-
temologically good relative to either the speaker or the hearer. That would
be persuasion without IP-justification. (Of course, it might still be a case
of subjective IP-justification, but it would not be objective IP-justification.)
In the final segment of this paper, I turn to the topic of good argumen-
tation, where I mean by ‘argumentation’ not a set of sentences or propo-
sitions, which is what I take an ‘argument’ to be, but an act of presenting
an argument to an audience or an interlocutor. Elsewhere (Goldman, 1994)
I have claimed that the principles or criteria of good argumentation do not
coincide with the criteria for good arguments understood in the logical
sense. For one thing, argumentation should be appropriately directed or
tailored to the speaker’s audience (or interlocutor), an element absent in
the criteria for a good argument in the logical sense. Here I want to discuss
the connection between good argumentation and IP-justification.
In what way should a piece of argumentation be tailored to its intended
audience? First of all, let us concentrate on cases in which the speaker
means to endorse or defend the presented argument, thereby excluding cases
in which one merely puts an argument forth for consideration, examina-
tion, or inquiry (cf. Meiland, 1989). Is it equally acceptable for the speaker
to endorse or present any argument that she herself accepts? I suggest as
a further constraint that a speaker should try to present arguments that have
ARGUMENTATION AND INTERPERSONAL JUSTIFICATION 163

the best prospect of being epistemologically good arguments relative to the


intended audience. This means that the speaker should try to select argu-
ments such that audience members will be justified in believing their
premises, and will be justified in believing that there is a ‘proper connec-
tion’ between the premises and the conclusion. Of course, a speaker cannot
guarantee that her attempts will succeed; in general, she may not know
what premises are justifiably believable by her various audience members.
But she should choose her arguments in such a way that, to the best of her
knowledge, they have as good a chance as possible of being epistemolog-
ically good relative to the audience (or as large a segment of the audience
as possible). In many cases, the premises to be used will be justifiably
believable prior to, and independently of, the speaker’s own argumenta-
tion. In other cases, the audience may become justified in believing one or
more of the premises in virtue of the speaker’s own authoritative assertion
of them.
Now if we add to the foregoing principle the requirement that the speaker
herself should be justified in believing the premises she endorses, and in
believing the propriety of the connection between her premises and con-
clusion, then we are in effect saying that a speaker should try to choose
arguments that enable her to IP-justify the conclusions to her audience.
(This will be IP-justification in the transmission sense.) Thus, IP-justifi-
cation may be identified as a guiding principle in the choice of arguments
to be directed at audiences.
Is IP-justification, then, the most fundamental aim of argumentation? I
think not. Elsewhere (Goldman, 1994) I have maintained that the core
purpose of argumentation is to persuade audiences or interlocutors of
truths. I do not mean to retract this suggestion in my current discussion of
IP-justification. Rather, I think there is a connection between IP-justifica-
tion and persuasion of truth. There is insufficient space here to spell out
the connection in detail, but I shall sketch it briefly. First, suppose we adopt
a reliabilist theory of P-justification (see Goldman, 1979, 1986). Then being
P-justified in believing a proposition implies that there is a substantial prob-
ability that this proposition is true. Thus, satisfying the requirements for
IP-justification, especially in the transmission sense, is positively related
to persuasion of truths. Second, suppose we assume that, typically, a person
who is justified in believing a proposition will actually believe it. For
example, if your perceptual experience P-justifies you in believing that a
cat is present, then typically you will believe that a cat is present. Then,
by the definition we have given of IP-justification, it follows that a speaker
who IP-justifies a conclusion to a hearer will make it likely (or at least raise
the probability) that the hearer will be persuaded of the conclusion. This
is because the hearer will be justified in believing the conjunction of the
premises, and justified in believing the propriety of the connection between
premises and conclusion. Thus, IP-justification is a means to the persua-
sion of truths (although, to be sure, not a fool-proof means). That, as I see
164 ALVIN I. GOLDMAN

it, is the most fundamental explanation of why argumentation should be


guided by the goal of IP-justification.

NOTE

1
Feldman’s full definition also specifies that an argument is good relative to a person only
if it’s not ‘defeated’ for him/her. I omit the topic of defeaters because it would require too
much space. As far as I can see, however, proper treatment of defeaters would not change
any of the points made in the rest of the paper.

REFERENCES

Feldman, R.: 1994, ‘Good Arguments’, in F. F. Schmitt (ed.), Socializing Epistemology:


The Social Dimensions of Knowledge, Rowman and Littlefield, Lanham, MD.
Goldman, A. I.: 1979, ‘What is Justified Belief?’, in G. S. Pappas (ed.), Justification and
Knowledge, Reidel, Dordrecht. Reprinted in: A. I. Goldman 1992, Liaisons: Philosophy
Meets the Cognitive and Social Sciences, MIT Press, Cambridge, MA.
Goldman, A. I.: 1986, Epistemology and Cognition, Harvard University Press, Cambridge,
MA.
Goldman, A. I.: 1994, ‘Argumentation and Social Epistemology’, Journal of Philosophy
91(1), 27–49.
Meiland, J.: 1989, ‘Argument as Inquiry and Argument as Persuasion’, Argumentation 3,
185–196.
Rorty, R.: 1979, Philosophy and the Mirror of Nature, Princeton University Press, Princeton.

You might also like