No Reason At All, Or Why Meaning Is Not Normative In The Reason-Implying Sense
Joanna Klimczyk
(University of Szczecin)
A popular and deemed innocent argument for the claim that meaning is normative
says that what an expression, or a word means provide the subject with some, though
not necessarily decisive, reason to employ it in certain ways (Whiting 2009: 537). The
same argument often takes the following form: that a word, or expression has
conditions for its correct use, speaks in favor, or counts in favor of using it in a certain
way. Among philosophers who believe that what this argument holds is true in quite
an embarrassingly obvious way are Daniel Whiting (2007; 2009) and Ralph
Wedgwood (2007) among others.
It is interesting to notice that the considered argument builds on the favoring
relation account of a normative reason, and its success in the debate over normativity
of meaning, depends on the success of the very theory of a normative reason that
underlies the argument in question. The philosopher who has brought the favoring
relation account of a reason into the philosophical attention, and who contributed
strongly to its upcoming fame was Thomas Scanlon in his seminal book What We
Owe To Each Other (1998).
In this paper I shall argue against the thesis that meaning is normative in the
reason-implying sense by means of attacking the Scanlonian account of a reason. The
idea behind my argumentative strategy is quite simple: if favoring relation account of
a reason is the correct theory of a normative reason, then its applicability to the area
of semantics should not be problematic. On the other hand, if we can prove that this
account of a reason fails tout court, the thesis that the lexical meaning of an
expression provides us with any reason to employ it in a certain way, will be
untenable. Briefly, if the favoring relation account of a reason is a good theory of a
normative reason, then, I contend, it is supposed to work when applied to any case at
hand, be it from moral, epistemic or semantic field.
It should be stressed that despite its initial appeal, the Scanlonian account of a
normative reason has been the object of heavy criticisms in the recent years. The
objections that have been raised against it revolve around the issues falling under
three categories: conceptual, ontological and functional. According to a complaint
belonging to the first category of the charges, it is far from clear why we would think
that the notion of a reason is a notion intrinsically normatively loaded. What
specifically stands in a need of explanation is why we would assume that whenever
one encounters a reason, a normative relation between the considerations taken to be
reason-giving and action, or attitude is to follow straightforwardly. The objection as it
stands does not deny that the notion of a reason can be normatively interpreted, what
it denies is rather that ‘a reason’ allows for one and only normative reading.
The second objection touches the problem regarding the metaphysical status of the
favoring relation, and recently it was meticulously discussed by Jonas Olson under the
label of ‘the extensional fallacy’ (Olson 2011). Olson observes that the problem with
this format of a normative reason account is that it generates theoretical commitments
it cannot give account of. Briefly, if a reason is to be cashed out in terms of a favoring
relation, then this favoring relation must be shown to be ‘real’, meaning it must refer
to something in the world. The problem with the Scanlonian account is not that he
understands a reason along the realistic lines, as something which as its extension has
any natural fact, but rather that he leaves us in the dark when it comes to
understanding what the relation of speaking in favor stands for. Finally, the last
1 objection takes as its target what is often regarded as perhaps the main advantage of
Scanlon’s theory of a reason, namely its intuitive attractiveness.
Scanlon seems to connect the intuitive appeal of his theory with the quietist
assumption with regard to what a reason is on which it is based. In the opening
statement of the first chapter of his book, he writes the following:
I will take the idea of a reason as primitive. Any attempt to explain what it is to be a
reason for something seems to me to lead back to the same idea: a consideration that
counts in favor of it. “Counts in favor how?” one might ask. “By providing a reason for it”
seems to be the only answer. So I will presuppose the idea of a reason, and presuppose
also that my readers are rational in the minimal but fundamental sense that I will presently
explain (Scanlon 1998:17).
In his further considerations Scanlon does not elaborate on the feature of
primitiveness that he ascribes to the notion of a reason, taking it probably as easily
decipherable, or at least sufficiently suggestive to speak for itself.
Since Scanlon’s account of a reason strikes me as neither intuitively compelling,
nor true, I shall offer my own interpretation of the primitiveness at stake. For the
purpose of my argument, I shall grant that the idea of a reason is primitive in the
relevant sense, if it is graspable as such by an ordinary agent who is rational in
Scanlon’s minimal sense of being rational. I shall also add two more “ifs” to my
working definition of the primitiveness we are after. A reason is primitive in the
relevant sense, if an ordinary agent who possess the idea of a reason in the primitive
sense, thinks of this reason as straightforwardly related to action or attitude as
Scanlon assumes, and if a consideration recognized as a reason does the job any
normative reason is supposed to do, whatever the details of one’s favorite theory of a
normative reason are. The last condition on a normative reason states simply that any
normative reason to deserve that name, be it construed primitively or not, must fulfil
three requirements: 1) explain why the agent has a normative reason for action, or
attitude; 2) be effectively action-guiding in the sense of playing a certain role in the
agent’s motivation, finally 3) offer a justification for action, or attitude that is of
genuine help when an agent tries to work out what it is that she ought to do in similar
circumstances in future.
I grant that the above-mentioned conditions on a normative reason are pretty
uncontroversial, insofar as we think of a normative reason through the function, or
functions it is to serve in our thinking. Hence, a normative reason, on any account, I
contend, must do well in any of the roles that are constitutive of it.
The problem with a Scanlonian account of a reason, as I shall further argue, and
quite aside from the conceptual and ontological worries it invites, which I will not
discuss in the paper due to the limited space, is that it conflicts with the idea of
primitiveness, which is supposed to be its cornerstone. The favoring relation account
of a reason turns out to be neither intuitively evident for a creature’s being minimally
rational in Scanlonian sense, nor workable. Its main flaw consists in its failure to
satisfy the constitutive constrains on a normative reason.
That’s for the general framework of my aim in the paper, now come the details.
That the Scanlonian account of a reason does not work as a theory of a reason will be
shown on the example of the normativist argument for the normativity of meaning. If
semantic norm speaks in favor of a certain linguistic behavior, hence is a normative
reason as the adherents of semantic normativism hold, then it must do the job of a
normative reason, and satisfy the constitutive requirements on a reason I have put
forward. The problem is that the bare fact about the meaning of an expression does
2 badly in the role of a reason. It neither explains why the subject applied the term
correctly, according with its meaning, when she did so, nor genuinely guides the
subject’s future uses of that term. Moreover, it is based on a controversial assumption
that reasons are factive. Suppose for the moment that my objections are in place, and
that meaning facts fail as candidates for normative reasons along the favoring relation
format of a reason that semantic normativists argue for. That need not entail a
devastating blow for the normativist position. Semantic norms may prove to be only
apparent reasons on one theory of a normative reason, and true normative reasons
according to some other. Anyway, we have quite a good deal of other theories of a
normative reason, and perhaps one of them fares well as an account of a semantic
normative reason. If that were the case, then our objection would turn out to be
misaddressed. It would not threaten semantic normativism as a tenable view, but at
best would demonstrate that the adherents of semantic normativism have chosen
badly on the market of theories of a reason. However, I shall show that things are not
so. Meaning norm fails in the role of reason also on the rival accounts of a reason as
defended by Broome (2004), or Hieronymi (2005, 2011, 2012). What is wrong with it
on these alternative accounts is that it proves to be either seriously explanatory
defective (the Broomean account), or gives rise to the well-known the wrong kind of
reason problem (Hieronymi). Taking all these things together, I shall conclude with
an obvious observation that, if some alleged reason does not pass the test for being a
reason as provided by the existing theories of a reason, then simply it is not a reason.
Or, at least those who think otherwise failed terribly in their task of making us believe
it is.
References:
Broome, J. (2004). ‘Reasons’. In: R. J. Wallace, P. Pettit, S. Scheffler, M. Smith
(eds.). Reason and Value. Themes from the Moral Philosophy of Joseph Raz. Oxford:
Clarendon Press, pp. 28-55.
Glüer, K., and Å. Wikforss. (2008). Against Normativity Again: Reply to Whiting,
MS, http://people.su.se/%7Ekgl/Reply%20to%20Whiting.pdf
Glüer, K., and Å. Wikforss. (2009). ‘The Normativity of Meaning and Content.’ The
Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/meaningnormativity/
Hattiangadi, A. 2009. ‘Some More Thoughts on Oughts and Thoughts: A Reply to
Daniel Whiting.’ Analysis. Volume 69, Number 1: 54-63.
Hieronymi, P. (2005). ‘The Wrong Kind of Reason,’ The Journal of Philosophy, 102
(9): 437-457.
Hieronymi, P. (2011). ‘Reasons For Action,’ Proceedings of the Aristotelian Society,
111: 407-427.
Hieronymi, P. (2012). The Wrong Kind of Reason, Revisited (MS).
Olson, J. (2009). ‘Reasons and The New Nonnaturalism’. In: S. Robertson (ed.).
Spheres of Reason. Oxford: Oxford University Press, pp. 164-182.
Scanlon, T. (1998). What We Owe To Each Other. Cambridge, Mass.: Harvard
University Press.
Wedgwood, R. (2007). The Nature of Normativity. Oxford: Oxford University Press.
Whiting, D. (2007). ‘The Normativity of Meaning Defended,’ Analysis 67: 130-140.
Whiting, D. (2009). ‘Is Meaning Fraught With Ought?’ Pacific Philosophical
Quarterly 90 (4): 535-555.
3