1

On The Inscrutability of Reference

“I want to begin with an epistemological accusation, or, rather, with an accusation against
epistemology. Epistemology is an expression of doubt in the power and the validity of
philosophical knowledge. It implies a division which undermines the possibility of knowledge.
Thinkers who devote themselves to epistemology seldom arrive at ontology. The path they
follow is not one which leads to reality.”

-Nicolas Berdyaev, The Destiny of Man; 1.

Stated briefly, the inscrutability of reference thesis holds that different kinds of analytical hypotheses can
be employed in the translation of an unknown language such that one and the same expression of that
language is interpreted in many incompatible ways; that the interpretations gotten thereby will impart
varying ontologies to the translated speaker; that they will be consistent with all observable behaviour
on the part of those being translated; and that “neither in the matter of termhood, nor in the matter of
reference, is there any sense to the question of there being a uniquely correct translation” (Gibson, 69-
70).

In this paper I will respond to elements of Quine’s arguments for the inscrutability of reference (ibid, 69-
79) through the device of a thought experiment. The experiment is that you and I are the natives that
Quine’s radical translator is translating (Quine, 121). By assumption, we are in the privileged position of
knowing that that translator does not know our language. Furthermore, we are well aware of the fact
that he thinks he faces certain intractable problems in the translation of our language. Using this
privileged perspective, we will identify a weakness of the inscrutability thesis.

As a methodological point, Quine’s own arguments invite us to consider ourselves as both the natives
and the radical translator: for inscrutability of reference is not only inter-cultural and inter-personal but
intra-personal, too, even for a monolingual (Alston; 58-59; c.f. Gibson, 73-74) 1. There is not, then, an
inconsistency in our thought experiment that derives from acting as both translator and translated.
Furthermore, while there might look to be something question begging about this approach, since it will
assume, at least as an opening stance, that reference is scrutable for competent speakers of the same
language, I propose it as a more natural, more plausible account of the facts of the matter than as a
refutation of Quine on grounds that he accepts.

That said, I will, during the course of my remarks, suggest that this response is in line with Quine’s
assumptions by running an argument against him that parallels the line Socrates runs against Protagoras
in the Theaetetus (Theaetetus, 171a-c). Still, my major concern is to provide not so much a refutation as
an attractive alternative to Quine’s account. This, too, is in keeping with Quine’s methods, for he
presents his own naturalistic-behaviouralistic (Gibson, 63-65), scientistic (lecture notes) program not as a
refutation of transcendental idealism, for instance, but as a project that will be more or less appealing to
those who already share or have a predominant tendency in favour of his assumptions 2. The method of
the thought experiment is itself a deliberate attempt to work with tools from Quine’s own toolbox (cf.
1
“...the inscrutability of reference can be brought even close to home than the neighbour’s case; we can apply it to
ourselves.” (Alston, 58-59).
 2

e.g. the “visiting Martian”; Quine, 139). Specifically, his radical translation scenario is itself a thought
experiment (Quine, 121: “I shall imagine...” etc.).

***

If I were to ask you to point to the index finger of your left hand using the index finger of your right hand,
you would, presumably, have no trouble fulfilling my request. If I were to then ask you to point to the
skin of the index finger on your left hand in the same way, your pointing might involve a tugging at the
outer surface and a sign or two to suggest that you don't mean the nail or the bone. You'd find a way to
signify the intended referent. If this isn’t especially obvious, imagine that the relationship between us is
that I am your surgeon and that if you do not signify the intended referent, I will not excise the intended
portion of your body. This consideration adds to the example the noble gravity of reality. Without this
modification, it might seem plausible at first glance that no amount of pointing, linguistic or otherwise,
could disambiguate sufficiently from a number of referents the intended one. Of course, since the
translator is working under a behaviourist theory of language (Gibson, 63-65), the surgery could in
principle be enough to disambiguate, since it would present a new set of observations to prompt for
stimulus meanings. But for us, right now, the surgery exists only in speech and not in deed, and because
the translator is having problems with our speech, he cannot, from within his framework, disambiguate
between what is the same under observation or by deed: index finger and skin of the same. The strange
thing is this: Quine says not only that a translator could not disambiguate the referents, but that there’s
nothing to be disambiguated (Gibson, 73)3.

In case my example is taken as saying something so obvious that Quine cannot have held otherwise –
that there is something to disambiguate and that we have the capacity to disambiguate, as evidenced, at
least, by our speech about deeds, let me point out that William Alston’s disbelief, which we share, at the
implication that there is no such disambiguation because no fact of the matter obtains is in no way
quelled by Quine’s response to Alston. In that response, Quine held that Alston is afraid of these
consequences because he is an “unregenerate heart” (footnote 2). Alston had written that “[it] is not
that the native really does mean ‘rabbit’ rather than ‘rabbit-stage’ by ‘gavagai’, though we can never find
this out. His meaning really is indeterminate as between the various alternatives mentioned” (Alston, 57;
emphasis added). Quine accepted Alston’s characterization as a fair representation of the thesis.
Similarly, Quine sanctions Gibson’s interpretation of his work with the praise that Gibson “understands
2
Concerning those assumptions, especially as contrasted, loosely but relevantly, with Socratic/Platonic
assumptions, say, consider the following statement of Quine’s: “...I philosophize from the vantage point only of our
own provincial conceptual scheme and scientific epoch, true; but I know no better” (Quine, 108). Also, as in his
response to William Alston’s Quine on Meaning (Alston, 75), which is the proximate cause of these reflections,
Quine refers to those whose arguments are couched in the very terms his thesis rejects as “unregenerate hearts,”
so Platonists, Aristotelians, Kantians and others may consider Quine’s own project the project of an unregenerate,
or “provincial” heart.

3
“...this is the point of the indeterminacy thesis[:] there are no grounds for declaring [a] particular system of
analytic hypotheses sacrosanct; and, since this is the case, there can be no useful sense to the question about what
[the foreigner’s expression] really means or really refers to. ‘Really’ thus queried is idle talk. As Quine insists, there
is no fact of the matter.” (Gibson, 73).
 3

my position so fully” as to give Quine great delight (Gibson, xi). The inscrutability thesis as I have
characterized it with the help of works by Gibson and Alston, and with reference to the thought
experiment, is Quine’s through and through. It’s this highly counter-intuitive and absurd thesis that is
the target of this paper4.

To return to our experiment, imagine now someone from a foreign land observing us in our finger-
identifying exercise. Imagine that he was convinced that when you were pointing to your finger,
although the interpretation that you were pointing out your finger was probable, it was impossible to
know for sure that you were not pointing out "finger-stage" or "finger-hood" or something other than
strictly "index-finger" or "skin-of-index-finger" as the case might be. Perhaps he’d think in the first
instance that you might have been pointing out the skin and that in the second instance you were
pointing out some quality of the skin. In short, imagine that he was convinced that he could never know;
that there was nothing to scrute. Furthermore, imagine that he was convinced that you could never
know, either, whether I had asked you to point at your finger or at some quality of your finger, for though
I uttered "finger," I might have just as well have said something substitutive that would have produced
the same observable result on your part. That is, if I had said point to a “portion-of-finger”, you would
have acted the same as when you pointed to your finger, for when you pointed in the latter case, you did
not make a grand sweeping gesture to indicate the whole finger, you merely pointed, and therefore to
merely a point, or portion, of finger5.

A few questions arise for us about this man. Why does he not seek to design further inquiries to sift out
the occasions on which I might plausibly be asking about "fingerhood" and those on which I am naturally
understood as indicating merely "the finger"? Reproducing the ‘gavagai’ example, to demonstrate my
meaning, we can disambiguate by devising this question: is a finger composed of undetached finger-
parts? Yes. Is an undetached finger part composed of fingers, amongst other components 6? No. We
have not forgotten that while such disambiguation is simple enough for us as competent speakers of the
same language, the translators are struggling to know how to get right the translation of the predicate
“is composed of.” But neither have we forgotten that their thesis is not merely that they don’t know but
that we don’t, either. It is true, as the translator would have it, that we can reinterpret for ourselves the
significance of “composed of” and perhaps come up with another answer, but it is more true, we think,
that there is a standard mode of operation when it comes to these things, that the translator has chosen
to ignore.

4
To the extent that Quine’s assumptions are part of the warp and woof of what L. Strauss called “the new political
science” in his “Epilogue” to Liberalism Ancient and Modern (Strauss, 203-223; cf. e.g. 210-211), the doctrine is not
only absurd but politically dangerous.

5
Even the case of a sweeping gesture, though, the translator tells us, is bound up with inscrutability, for the
translator’s hypothesis could allow the sweeping gesture to indicate a collection of undetached finger parts
(Gibson, 71).

6
In case the answer comes back from Quine that it depends what we mean by “finger,” I respond that our
experimental selves “mean” that differentiated part of the hand of which there are four, not counting the thumb;
they mean the intuitively obvious finger; the pre-scientific finger, as it were; the finger of common sense.
 4

Put it this way: the translator knows that he is translating the language of human beings, not of alien
creatures from another genus or species. As himself human, he knows, via his extensive familiarity with
myriad other foreign-language speakers around the world with whom he communicates, that the notion,
for example, of the essence of a thing, it's "x-hood" is not something that is commonly indicated by
pointing out objects of sense. In fact, he knows that the “x-hood” of a thing is not given empirically but
only if at all provoked empirically. He needn’t have read Kant to know this. We cannot understand why
he accepts as an “analytic hypothesis” something as absurd as that we might intend “x-hood” and not
“x” when we are speaking in our daily rounds (Quine, 143).

We hear him speaking of a scientifically precise and technical language that must not contain intensional
contexts and at the same time of a commitment to a naturalistic-behaviouristic conception of language.
But if these two desires result in such a consequence, mustn’t the translator rather abandon or modify
the desires than accept the consequences thereof 7? We suggest that we see in Quine’s thesis a reductio
of the argument, and thus of those fundamental desires or commitments (Gibson, 76) 8. Quine rejected
as a reductio those proofs that call into doubt the method of proof by reductio (Quine, 180). Are we not
justified by the same sort of reasoning in calling a reductio the argument that casts doubt on whether
competent, healthy, intelligent human speakers intends to signify by a word a thing or the undetached
parts of that thing?

To the extent that Quine believes that there’s no fact of the matter about what is referred to even for a
speaker himself in his own language, Quine is admitting, as we see it, that he doesn’t know what he is
talking about. There’s no fact of the matter of what he’s referring to and all of his utterances are subject
to proxy-functional shifts (Quine, 115), to radical reinterpretations. We are at liberty, we think, if we
choose to accept his utterances as plausible, to understand by his phrase “the inscrutability of
reference”, “the scrutability of reference”, or “the inscrutability-of-reference-for-Quine”, or what we will.
Since we cannot know what he is pointing out by those words, since he himself says he pointing out
nothing at all, we will suppose him to be pointing out what to us appears most plausible 9, that we can
cunningly readjust our translations of what he writes so as to compensate for the switch to our preferred
ontology. Let there be no fact of the matter as to what Quine is on about. We shall take him at his word
that there’s not.
7
I understand that Quine eventually came to accept intensionality in the language best fitted to work with the best
science of his time, but I have not heard that Quine abandoned the NB conception of language, which, as seen in
the following footnote, entails the inscrutability thesis. Thus, not only extensionality but the NB conception of
language must be dropped if the translators want to be true to how our language – the language of us in the
thought experiment – operates as a matter of fact.

8
“...Quine’s doctrine of ontological relativity..rests, of course, upon the NB [naturalistic-behavioristic] conception of
language: for the doctrine of relativity is really nothing more than a generalization of the doctrine of of the
inscrutability of reference, which is itself a consequence of the NB conception of language.” (Gibson, 76).

9
Quine writes in various places that the conceptual scheme we adopt is relative to our “various interests and
purposes” (e.g. Quine, 192). Our interests and purposes, granting Quine his argument, for a moment, involve
radically reinterpreting radical reinterpretation for the purpose of demonstrating the common sense soundness of
the scrutability of reference thesis.
 5

But wait. When I asked you to point to your finger, you did. When I asked you to indicate the skin of
your finger, you did. If I ask you further whether in the first instance you were pointing out finger-hood,
and I tell you that by finger-hood I mean the essential and general being that makes a particular thing a
particular thing of a particular kind, namely: a finger, you’d respond – indeed, you have responded, let’s
say - that you were not. For me to pose the question to you and for you to answer it there must be an
understanding between us of the notions of particularity, generality, being, essence, thing-hood, and so
on, and the claim that there is enough of an understanding between us on those words is something
that our foreigners deny, in the peculiar way in which they enjoy denying what to us is obvious. But we
have found no good reason to take their claim seriously, for as applied to their own utterances it gave us
license to suppose the contrary: that I know and you know whether the words we are using are or are
not referring to an object, to a property of an object, to the essence of an object, and so on. Moreover, I
know and you know that there can be disagreement or misunderstanding on these matters, and that we
have the linguistic wherewithal to resolve those difficulties or if not to resolve them to allow them to
become problems for us more or less capable of resolution.

From our privileged perspective it surely seems like something about this translator is amiss, that his
understanding of speech and understanding is deliberately confused. After all, his inscrutability thesis 10
arose from a perverse contrivance: he convinced himself that he could convince himself that we meant
just about anything if he reinterpreted just about everything we said. We do not deny that a person can
spin his own consistent delusion and justify what he will about what he will. We deny that this is natural,
that it is healthy, and that it is what fully functioning members of our speech community do. For us it
does not follow from the more or less consistent delusions of the madmen of our community that all
speech is fundamentally madness, that we’re all equally delusional.

Our position on this question differs from the translator’s position in a way that I’d like to bring out more
clearly. For us, there are healthy and unhealthy understandings, and there is a notion of the nature of
understanding and of speech in our community that grounds our rational discourse 11. We notice that in
the translator’s lands, as previously mentioned, people do not typically talk about essences-of-things:
they do not order coffee-hood in their cafes, but coffee; they do not wear jacketness but jackets. In the
contexts of their practice of advertisement, we notice that they sell sexiness, not sex. In the academic
context, we hear them saying that the suffixed ‘-ness’ denotes a quality that may or may not be
considered as a universal object. We notice them making such statements as “brittleness is the quality a
thing has of being easily broken.” We might hear, if we suffer to put ourselves in the context most likely
to facilitate the utterance, that “dogness” is indistinguishable from “dog” or from “dog-part” in any
community of speakers and listeners including the community of one. This is, of course, what we hear
when we listen to our unusual translator discussing the inscrutability thesis.

10
I recognize that I have by now collapsed “the translator” and “Quine” or that I’m using those terms
interchangeably. Here’s the fact of the matter: I mean Quine in every instance.

11
Needless to say, that understanding is not the NB concept of language simpliciter. For the purposes of this paper,
our understanding is roughly “transcendental” at least in part.
 6

In each case, the context for us was decisive, and where it would not have been, there would have been
recourse to further inquiry to settle the score or to articulate a problem, as I have said. What’s
conspicuously missing from our perspective in the translator’s theses about our language acts is an
appreciation of natural context (of nature as such, in fact; of physis as distinct from nomos). They
speculated not that I might mean ‘finger-hood’ and not ‘finger’ within the context of a seminar on the
philosophy of language, but that I might mean it when I am having a ring sized or a cut treated. From
our perspective, this is a serious and conspicuous defect in the translator’s background assumptions.

Socrates said of Protagoras’ “Truth” that on its own account it could very well be false, for if Socrates and
Theatetus thought it was, then, for them, it was, and if everyone else in the community thought so too,
then it was only truth for Protagoras, and falsehood for all others (Theaetetus, 171a-c). He got
Theaetetus out of error with the thought that you do not trust anyone to make a good prediction, but
him who knows; and you do not think that every man knows, but he who makes a good prediction, in
certain cases, at least. Socrates, I think, showed us how to take the thesis of our translators, who claim
that no man understands what another is saying, nor what he himself is saying. We shall not buy a Buick
from such translators. And we shall be very, very careful indeed when considering what of theirs we
have hastily purchased heretofore, for those purchases were not of automobiles but of world-views that
concern the most important questions, the most important matters.

“Truth is immanent, and there is no higher,” says Quine (Quine, 246) from within his own “background
language” (Gibson, 76). But Plato’s Socrates spoke of those who, undergoing the philosophical
education, “...lift up the radiant light of their souls to what itself provides light for everything” (Republic,
540a). The philosophical man “attempts to grasp with respect to each thing itself what the being of it is”
(ibid, 533b) “until he grasps the good itself with understanding itself” and “reaches the end of the
intelligible” (ibid, 523a-b). That is, the original promise of philosophy is the promise of life lived in
accordance with a transcendental truth and with reality and being, with the turning of the soul away
from utter immanence (and opinion12) toward a truth that is “unhypothetical” or simply true (ibid, 513b);
i.e. to a principle, or first thing, that is not itself relative to a set of hypothesis, that does not arise from
“within a theory” (Quine, 246)13. Quine also says of himself and his own ontological commitments that
he, “qua lay physicist,” believes in physical objects and “not in Homer’s gods” (ibid, 52). Furthermore, he
considers it “a scientific error to believe otherwise” (ibid). At the same time, he says that physical
objects and the gods are both, from an epistemic perspective, merely “cultural posits” and “myths”
(ibid). He says that the physical object myth has been more successful than other myths because it
succeeds in “working a manageable structure into the flux of experience” (ibid). And I remind the reader
that Quine writes from a “provincial conceptual scheme” because he knows no better (note 2). Putting
this all together, we suggest the following. Whether or not one grants Quine’s arguments – that is, from
12
“What? Haven’t you noticed that opinions without knowledge are shameful and ugly things?” (Republic, 506c).

13
Cf. the beginning of Strauss, “What Can We Learn From Political Theory,” for the classical as opposed to the
modern notion of theory and in general for clarification on my attempt in this paper to move from the latter to the
former sense of the term, from Quine to Plato, from the “background languages” of his relativistic theorizing to the
“coherent reflection” on “the right standards of judgment” and the notion of “the” truth simply in Plato. For the
connection of inscrutability to political matters, cf. note 4.
 7

a natural and from a contrived position both – what matters is what project one is undertaking, what
“structure” will guide, give significance to, and call forth actions and aims in one’s “experience” of living,
for if one places the love of wisdom and the philosophical life above all other lives, above even the
“provincial conceptual scheme” of one’s “scientific epoch”, then the “myths” and “cultural posits” that
one will select in orienting one’s life will be those myths that allow for the possibility of a transcendental
truth (“how greedy for images I am”, said Socrates, for those images served his psychogogic project;
Republic 488a), those that do not cut off the possibility of philosophy from the start with such
dogmatisms as that “truth is immanent, and there is no higher.” This kind of dogmatizing, we have seen,
has led Quine and his radical translators into a coal-pit (Alston, 59). Whether we follow him into that pit
or whether we remain unregenerate and philosophical men is a choice that ours to make.

WORKS CITED

Alston; Quine on Meaning; in The Philosophy of W.V. Quine, Library of Living Philosophers Vol. XVIII;
edited by Hahn and Schlipp; Open Court Publishing Company; 1986.

Berdyaev; The Destiny of Man; Harper and Brothers; 1960.

Gibson; The Philosophy of W.V. Quine; University Presses of Florida; 1974.

Plato; Complete Works; edited by Cooper and Hutchinson; Hackett Publishing Company; 1997.

Quine; Quintessence: Basic Readings from the Philosophy of W.V. Quine; edited by Gibson;
Belknap/Harvard University Press; 2004.

Strauss; Liberalism Ancient and Modern; Cornell University Press; 1968.

Strauss; What Can We Learn From Political Theory; accessed online: