Confirmation A4

pdf version of the entry
Confirmation
https://plato.stanford.edu/archives/spr2021/entries/confirmation/ Confirmation
from the Spring 2021 Edition of the First published Thu May 30, 2013; substantive revision Tue Jan 28, 2020
Stanford Encyclopedia Human cognition and behavior heavily relies on the notion that evidence
(data, premises) can affect the credibility of hypotheses (theories,
of Philosophy conclusions). This general idea seems to underlie sound and effective
inferential practices in all sorts of domains, from everyday reasoning up to
the frontiers of science. Yet it is also clear that, even with extensive and
truthful evidence available, drawing a mistaken conclusion is more than a
mere possibility. For painfully tangible examples, one only has to consider
missed medical diagnoses (see Winters et al. 2012) or judicial errors (see
Edward N. Zalta Uri Nodelman Colin Allen R. Lanier Anderson
Liebman et al. 2000). The Scottish philosopher David Hume (1711–1776)
Principal Editor Senior Editor Associate Editor Faculty Sponsor
is usually credited for having disclosed the theoretical roots of these
Editorial Board
https://plato.stanford.edu/board.html considerations in a particularly transparent way (see Howson 2000, Lange
2011, and Varzi 2008). In most cases of interest, Hume pointed out, many
Library of Congress Catalog Data
ISSN: 1095-5054
alternative candidate hypotheses remain logically compatible with all the
relevant information at one’s disposal, so that none of the former can be
Notice: This PDF version was distributed by request to mem- singled out by the latter with full certainty. Thus, under usual
bers of the Friends of the SEP Society and by courtesy to SEP circumstances, reasoning from evidence must remain fallible.
content contributors. It is solely for their fair use. Unauthorized
distribution is prohibited. To learn how to join the Friends of the This fundamental insight has been the source of a lasting theoretical
SEP Society and obtain authorized PDF versions of SEP entries, challenge: if amenable to analysis, the role of evidence as supporting (or
please visit https://leibniz.stanford.edu/friends/ . infirming) hypotheses has to be grasped by more nuanced tools than plain
logical entailment. As emphasized in a joke attributed to American
Stanford Encyclopedia of Philosophy
Copyright c 2021 by the publisher philosopher Morris Raphael Cohen (1880–1947), logic textbooks had to
The Metaphysics Research Lab be divided in two parts: in the first part, on deductive logic, unwarranted
Center for the Study of Language and Information
forms of inference (deductive fallacies) are exposed; in the second part, on
Stanford University, Stanford, CA 94305
inductive logic, they are endorsed (see Meehl 1990, 110). In contemporary
Confirmation
Copyright c 2021 by the author philosophy, confirmation theory can be roughly described as the area
Vincenzo Crupi where efforts have been made to take up the challenge of defining
All rights reserved. plausible models of non-deductive reasoning. Its central technical term—
Copyright policy: https://leibniz.stanford.edu/friends/info/copyright/
1
Confirmation Vincenzo Crupi
confirmation—has often been used more or less interchangeably with 2. Hypothetico-deductivism

“evidential support”, “inductive strength”, and the like. Here we will 2.1 HD vs. Hempelian confirmation
generally comply with this liberal usage (although more subtle conceptual 2.2 Back to black (ravens)
and terminological distinctions are sometimes drawn). 2.3 Underdetermination and the Duhemian challenge
2.4 The extended HD menu
Confirmation theory has proven a rather difficult endeavour. In principle, it 3. Bayesian confirmation theories
would aim at providing understanding and guidance for tasks such as 3.1 Probabilistic confirmation as firmness
diagnosis, prediction, and learning in virtually any area of inquiry. Yet 3.2 Strengths and infirmities of firmness
popular accounts of confirmation have often been taken to run into 3.3 Probabilistic relevance confirmation
troubles even when faced with toy philosophical examples. Be that as it 3.4 Differences, ratios, and partial entailment
may, there is at least one real-world kind of activity which has remained a 3.5 New evidence, old evidence, and total evidence
prevalent target and benchmark, i.e., scientific reasoning, and especially 3.6 Paradoxes probabilified and other elucidations
key episodes from the history of modern and contemporary natural Bibliography
science. The motivation for this is easily figured out. Mature sciences Academic Tools
seem to have been uniquely effective in relying on observed evidence to Other Internet Resources
establish extremely general, powerful, and sophisticated theories. Indeed, Related Entries
being capable of receiving genuine support from empirical evidence is
itself a very distinctive trait of scientific hypotheses as compared to other
kinds of statements. A philosophical characterization of what science is 1. Confirmation by instances
would then seem to require an understanding of the logic of confirmation.
And so, traditionally, confirmation theory has come to be a central concern In a seminal essay on induction, Jean Nicod (1924) offered the following
of philosophers of science. important remark:
In the following, major approaches to confirmation theory are overviewed Consider the formula or the law: F entails G. How can a particular
according to a classification that is relatively standard (see Earman and proposition, or more briefly, a fact affect its probability? If this fact
Salmon 1992; Norton 2005): confirmation by instances (Section 1), consists of the presence of G in a case of F, it is favourable to the
hypothetico-deductivism and its variants (Section 2), and probabilistic law […]; on the contrary, if it consists of the absence of G in a
(Bayesian) approaches (Section 3). case of F, it is unfavourable to this law. (219, notation slightly
adapted)
1. Confirmation by instances
1.1 Hempel’s theory Nicod’s work was an influential source for Carl Gustav Hempel’s (1943,
1.2 Two paradoxes and other difficulties 1945) early studies in the logic of confirmation. In Hempel’s view, the key
2 Stanford Encyclopedia of Philosophy Spring 2021 Edition 3

valid message of Nicod’s statement is that the observation report that an The notion of the e -development of hypothesis h can be given an entirely
object a displays properties F and G (e.g., that a is a swan and is white) general and precise definition, but we’ll not need this level of detail here.
confirms the universal hypothesis that all F-objects are G -objects (namely, Suffice it to say that the e -development of a universally quantified material
that all swans are white). Apparently, it is by means of this kind of conditional ∀x(Fx → Gx) is just as expected, that is: Fa → Ga in case
confirmation by instances that one can obtain supporting evidence for I(e) = {a} ; (Fa → Ga) ∧ (Fb → Gb) in case I(e) = {a, b}, and so on.
statements such as “sodium salts burn yellow”, “wolves live in a pack”, or Following Hempel, we will take universally quantified material
“planets move in elliptical orbits” (also see Russell 1912, Ch. 6). We will conditionals as canonical logical representations of relevant hypotheses.
now see the essential features of Hempel’s analysis of confirmation. So, for instance, we will count a statement of the form ∀x(Fx → Gx) as an
adequate rendition of, say, “all pieces of copper conduct electricity”.
1.1 Hempel’s theory
In Hempel’s theory, evidence statement e is said to confirm hypothesis h
Hempel’s theory addresses the non-deductive relation of confirmation just in case it entails, not h in its full extension, but suitable instantiations
between evidence and hypothesis, but relies thoroughly on standard logic of h. The technical notion of the e -development of h is devised to identify
for its full technical formulation. As a consequence, it also goes beyond precisely those relevant instantiations, that is, the consequences of h as
Nicod’s idea in terms of clarity and rigor. restricted to the individuals involved in e. More precisely, Hempelian
confirmation can be defined as follows:
Let L be the set of the closed sentences of a first-order logical language L
(finite, for simplicity) and consider h, e ∈ L . Also let e , the evidence Hempelian confirmation
statement, be consistent and contain individual constants only (no For any h, e ∈ L such that e is consistent and contains individual
quantifier), and let I(e) be the set of all constants occurring (non- constants only (no quantifier):
vacuously) in e . So, for example, if e = Qa ∧ Ra, then I(e) = {a} , and if
i. evidence e directly Hempel-confirms hypothesis h if and only if
e = Qa ∧ Qb , then I(e) = {a, b} . (The non-vacuity clause is meant to
e ⊨ deve (h) ; e Hempel-confirms h if and only if, for some s ∈ L,
ensure that if sentence e happens to be, say, Qa ∧ Qb ∧ (Rc ∨ ¬Rc), then
e ⊨ deve (s) and s ⊨ h ;
I(e) still is {a, b} , for e does not really state anything non-trivial about the
ii. evidence e directly Hempel-disconfirms hypothesis h if and only
individual denoted by c . See Sprenger 2011a, 241–242.) Hempel’s theory
if e ⊨ deve (¬h); e Hempel-disconfirms h if and only if, for some
relies on the technical construct of the development of hypothesis h for
s ∈ L, e ⊨ deve (s) and s ⊨ ¬h;
evidence e , or the e -development of h, indicated by deve (h) . Intuitively,
iii. evidence e is Hempel-neutral for hypothesis h otherwise.
deve (h) is all that (and only what) h says once restricted to the individuals
mentioned (non-vacuously) in e , i.e., exactly those denoted by the In each of clauses (i) and (ii), Hempelian confirmation (disconfirmation,
elements of I(e). respectively) is a generalization of direct Hempelian confirmation

(disconfirmation). To retrieve the latter as a special case of the former, one disconfirms that all swans are white;
only has to posit s = h (¬h, respectively, for disconfirmation). (the observation report of) a non-white swan also Hempel-
disconfirms that a further swan will be white.
By direct Hempelian confirmation, evidence statement e that, say, object a
is a white swan, swan(a) ∧ white(a), confirms hypothesis h that all swans 1.2 Two paradoxes and other difficulties
are white, ∀x(swan(x) → white(x)), because the former entails the e-
development of the latter, that is, swan(a) → white(a) . This is a desired The ravens paradox (Hempel 1937, 1945). Consider the following
result, according to Hempel’s reading of Nicod. By (indirect) Hempelian statements:
confirmation, moreover, swan(a) ∧ white(a) also confirms that a
particular further object b will be white, if it’s a swan, i.e., ( h) ∀x(raven(x) → black(x)), i.e., all ravens are black;
swan(b) → white(b) (to see this, just set s = ∀x(swan(x) → white(x))). (e) raven(a) ∧ black(a), i.e., a is a black raven;
(e∗ ) ¬black(a∗ ) ∧ ¬raven(a∗ ), i.e., a∗ is a non-black non-raven (say, a
The second possibility considered by Nicod (“the absence of G in a case green apple).
of F ”) can be accounted for by Hempelian disconfirmation. For the
evidence statement e that a is a non-white swan—swan(a) ∧ ¬white(a) — Is hypothesis h confirmed by e and e∗ alike? That is, is the claim that all
entails (in fact, is identical to) the e -development of the hypothesis that ravens are black equally confirmed by the observation of a black raven and
there exist non-white swans—∃x(swan(x) ∧ ¬white(x)) —which in turn is by the observation of a non-black non-raven (e.g., a green apple)? One
just the negation of ∀x(swan(x) → white(x)). So the latter is disconfirmed would want to say no, but Hempel’s theory is unable to draw this
by the evidence in this case. And finally, e = swan(a) ∧ ¬white(a) also distinction. Let’s see why.
Hempel-disconfirms that a particular further object b will be white, if it’s a
As we know, e (directly) Hempel-confirms h, according to Hempel’s
swan, i.e., swan(b) → white(b), because the negation of the latter,
reconstruction of Nicod. By the same token, e∗ (directly) Hempel-
swan(b) ∧ ¬white(b) , is entailed by s = ∀x(swan(x) ∧ ¬white(x)) and
e ⊨ deve (s).
confirms the hypothesis that all non-black objects are non-ravens, i.e.,
h∗ = ∀x(¬black(x) → ¬raven(x)). But h∗ ⊨ h (h and h∗ are just logically
So, to sum up, we have four illustrations of how Hempel’s theory equivalent). So, e∗ (the observation report of a non-black non-raven), like
articulates Nicod’s basic ideas, to wit: e (black raven), does (indirectly) Hempel-confirm h (all ravens are black).
Indeed, as ¬raven(a) entails raven(a) → black(a), it can be shown that h
(the observation report of) a white swan (directly) Hempel-confirms is (directly) Hempel-confirmed by the observation of any object that is not
that all swans are white; a raven (an apple, a cat, a shoe), apparently disclosing puzzling “prospects
(the observation report of) a white swan also Hempel-confirms that a for indoor ornithology” (Goodman 1955, 71).
further swan will be white;
(the observation report of) a non-white swan (directly) Hempel- Blite (Goodman 1955). Consider the peculiar predicate “blite ”, defined as
follows: an object is blite just in case (i) it is black if examined at some

moment t up to some future time T (say, the next expected appearance of discussion (Rinard 2014 is a recent exception). After all, for all we know,
Halley’s comet, in 2061) and (ii) it is white if examined afterwards. So we it is a perfectly “natural” feature of a token of the “natural kind” water that
posit blite(x) ≡ (ext≤T (x) → black(x)) ∧ (¬ext≤T (x) → white(x)). Now it is found in one physical state for temperatures below 0 degrees Celsius
consider the following statements: and in an entirely different state for temperatures above that threshold. So
why should the time threshold T in blite or blurple be a reason to dismiss
( h) ∀x(raven(x) → black(x)), i.e., all ravens are black; those predicates? (The water example comes from Howson 2000, 31–32.
(h∗ ) ∀x(raven(x) → blite(x)) , i.e., all ravens are blite; See Schwartz 2011, 399 ff., for a more general assessment of this issue.)
(e) e = raven(a) ∧ ext≤T (a) ∧ black(a), i.e., a is a raven observed no
later than T and it is black. The above, widely known “paradoxes” then suggest that Hempel’s
analysis of confirmation is too liberal: it sanctions the existence of
Does e confirm hypotheses h and h∗ alike? That is, does the observation of confirmation relations that are intuitively very unsound (see Earman and
a black raven before T confirms equally the claim that all ravens are black Salmon 1992, 54, and Sprenger 2011a, 243, for more on this). Yet the
as the claim that all ravens are blite? Here again, one would want to say Hempelian notion of confirmation turns out to be very restrictive, too, on
no, but Hempel’s theory is unable to draw the distinction. For one can other accounts. For suppose that hypothesis h and evidence e do not share
check that the e -developments of h and h∗ are both entailed by e. Thus, e any piece of non-logical vocabulary. h might be, say, Newton’s law of
(the report of a raven examined no later than T and found to be black) universal gravitation (connecting force, distances and masses), while e
does Hempel-confirm h∗ (all ravens are blite) just as it confirms h (all could be the description of certain spots on a telescopic image.
ravens are black). Moreover, e also Hempel-confirms the statement that a Throughout modern physics, significant relations of confirmation and
raven will be white if examined after T , because this is a logical disconfirmation were taken to obtain between statements like these.
consequence of h∗ (which is directly Hempel-confirmed by e). And finally, Indeed, telescopic sightings have been crucial evidence for Newton’s law
suppose that as applied to celestial bodies. However, as their non-logical vocabularies
blurple(x) ≡ (ext≤T (x) → black(x)) ∧ (¬ext≤T (x) → purple(x)). We then are disjoint, e and h must simply be logically independent, and so must be
have that the very same evidence statement e Hempel-confirms the e and deve (h) (with very minor caveats, this follows from Craig’s so-
hypothesis that all ravens are blurple, and thus also its implication that a called interpolation theorem, see Craig 1957). In such circumstances, there
raven will be purple if examined after T ! can be nothing but Hempel-neutrality between evidence and hypothesis.
So Hempel’s original theory seems to lack the resources to capture a key
A seemingly obvious idea, here, is that there must be something inherently
feature of inductive inference in science as well as in several other
wrong with predicates such as blite or blurple (and perhaps non-raven
domains, i.e., the kind of “vertical” relationships of confirmation (and
and non-black, too) and thus a principled way to rule them out as
disconfirmation) between the description of observed phenomena and
“unnatural”. Then one could restrict confirmation theory accordingly, i.e.,
hypotheses concerning underlying structures, causes, and processes.
to “natural kinds” only (see, e.g., Quine 1970). Yet this point turns out be
very difficult to pursue coherently and it has not borne much fruit in this

To overcome the latter difficulty, Clark Glymour (1980a) embedded a One can see that in the above definition the auxiliary assumptions in k are
refined version of Hempelian confirmation by instances in his analysis of the e -development of further closed constant-free hypotheses (in fact,
scientific reasoning. In Glymour’s revision, hypothesis h is confirmed by equations as applied to specific measured values, in typical examples from
some evidence e even if appropriate auxiliary hypotheses and assumptions Glymour 1980a), where such hypotheses are meant to be conjoined in a
must be involved for e to entail the relevant instances of h. This important single statement (α) for convenience. This implies that the only terms
theoretical move turns confirmation into a three-place relation concerning occurring (non-vacuously) in k are individual constants already occurring
the evidence, the target hypothesis, and (a conjunction of) auxiliaries. (non-vacuously) in e. For an empty α (that is, tautologous: α = ⊤), k must
Originally, Glymour presented his sophisticated neo-Hempelian approach be empty too, and the original (restricted) definition of Hempelian
in stark contrast with the competing traditional view of so-called confirmation applies. As for the proviso that α ⊭ h, it rules out undesired
hypothetico-deductivism (HD). Despite his explicit intentions, however, cases of circularity—akin to so-called “macho” bootstrap confirmation, as
several commentators have pointed out that, partly because of the due discussed in Earman and Glymour 1988 (for more on Glymour’s theory
recognition of the role of auxiliary assumptions, Glymour’s proposal and and its developments, see Douven and Meijs 2006, and references
HD end up being plagued by similar difficulties (see, e.g., Horwich 1983, therein).
Woodward 1983, and Worrall 1982). In the next section, we will discuss
the HD framework for confirmation and also compare it with Hempelian 2. Hypothetico-deductivism
confirmation. It will thus be convenient to have a suitable extended
definition of the latter, following the remarks above. Here is one that The central idea of hypothetico-deductive (HD) confirmation can be
serves our purposes: roughly described as “deduction-in-reverse”: evidence is said to confirm a
hypothesis in case the latter, while not entailed by the former, is able to
Hempelian confirmation (extended) entail it, with the help of suitable auxiliary hypotheses and assumptions.
For any h, e, k ∈ L such that e contains individual constants only (no The basic version (sometimes labelled “naïve”) of the HD notion of
quantifier), k = deve (α) for some α ∈ L containing quantifiers only confirmation can be spelled out thus:
(no individual constant) and such that α ⊭ h, and e ∧ k is consistent:
HD-confirmation
i. e directly Hempel-confirms h relative to k if and only if For any h, e, k ∈ L such that h ∧ k is consistent:
e ∧ k ⊨ deve (h); e Hempel-confirms h relative to k if and only if,
for some s ∈ L, e ∧ k ⊨ deve (s) and s ∧ k ⊨ h; i. e HD-confirms h relative to k if and only if h ∧ k ⊨ e and k ⊭ e ;
ii. e directly Hempel-disconfirms h relative to k if and only if ii. e HD-disconfirms h relative to k if and only if h ∧ k ⊨ ¬e, and
e ∧ k ⊨ deve (¬h); e Hempel-disconfirms h relative to k if and k ⊭ ¬e;
only if, for some s ∈ L, e ∧ k ⊨ deve (s)a and s ∧ k ⊨ ¬h; iii. e is HD-neutral for hypothesis h relative to k otherwise.
iii. e is Hempel-neutral for h relative to k otherwise.

Note that clause (ii) above represents HD-disconfirmation as plain logical It is rather common for a theory of ampliative (non-deductive) reasoning
inconsistency of the target hypothesis with the data (given the auxiliaries) to retain classical logical entailment as a special case (a feature sometimes
(see Hempel 1945, 98). called “super-classicality”; see Strasser and Antonelli 2019). That’s
essentially what (EC) implies for confirmation. Now given appropriate e ,
2.1 HD vs. Hempelian confirmation h and k, if e ∧ k entails h , we readily get that e Hempel-confirms h
relative to k in two simple steps. First, given that e and k are both
HD-confirmation and Hempelian confirmation convey different intuitions quantifier-free, deve (e ∧ k) = e ∧ k according to Hempel’s full definition
(see Huber 2008a for an original analysis). They are, in fact, distinct and of dev (see Hempel 1943, 131). Then it trivially follows that
strictly incompatible notions. This will be effectively illustrated by the e ∧ k ⊨ deve (e ∧ k), so e ∧ k is (directly) Hempel-confirmed and its
consideration of the following conditions. logical consequence h is likewise confirmed (indirectly). Logical
entailment is thus retained as an instance of Hempelian confirmation in a
Entailment condition (EC)
fairly straightforward way. HD-confirmation, on the contrary, does not
For any h, e, k ∈ L , if e ∧ k is consistent, e ∧ k ⊨ h and k ⊭ h , then e
fulfil (EC). Here is one odd example (see Sprenger 2011a, 234). With
confirms h relative to k.
k = ⊤ , just let e be the observation report that object a is a black swan,
Confirmation complementarity (CC) swan(a) ∧ black(a), and h be the hypothesis that black swans exist,
For any h, e, k ∈ L , e confirms h relative to k if and only if e ∃x(swan(x) ∧ black(x)). Evidence e verifies h conclusively, and yet it
disconfirms ¬h relative to k. does not HD-confirm it, simply because h ⊭ e. So the observation of a
black swan turns out to be HD-neutral for the hypothesis that black swans
Special consequence condition (SCC) exist! The same example shows how HD-confirmation violates (CC), too.
For any h, e, k ∈ L , if e confirms h relative to k and h ∧ k ⊨ h∗ , then In fact, while HD-neutral for h , e HD-disconfirms its negation ¬h that no
e confirms h∗ relative to k. swan is black, ∀x(swan(x) → ¬black(x)), because the latter is obviously
inconsistent with (refuted by) e .
On the implicit proviso that k is empty (that is, tautologous: k = ⊤ ),
Hempel (1943, 1945) himself had put forward (EC) and (SCC) as The violation of (EC) and (CC) by HD-confirmation is arguably a reason
compelling adequacy conditions for any theory of confirmation, and for concern, for those conditions seem highly plausible. The special
devised his own proposal accordingly. As for (CC), he took it as a plain consequence condition (SCC), on the other hand, deserves separate and
definitional truth (1943, 127). Moreover, Hempelian confirmation careful consideration. As we will see later on, (SCC) is a strong constraint,
(extended) satisfies all conditions above (of course, for arguments h, e and and far from sacrosanct. For now, let us point out one major philosophical
k for which it is defined). HD-confirmation, on the contrary, violates all of motivation in its favor. (SCC) has often been invoked as a means to ensure
them. Let us briefly discuss each one in turn. the fulfilment of the following condition (see, e.g., Hesse 1975, 88;
Horwich 1983, 57):

Predictive inference condition (PIC) auxiliary assumption k = raven(a) is made—and a is checked for color
For any e, k ∈ L, if e confirms ∀x(Fx → Gx) relative to k , then e and found to be black, then, yes, the latter evidence, black(a) , HD-
confirms F(a) → G(a) relative to k. confirms that all ravens are black (h) relative to k . By the same token,
¬black(a) HD-disconfirms h relative to the same assumption
In fact, (PIC) readily follows from (SCC) and so it is satisfied by Hempel’s k = raven(a) . And, again, this is as it should be, in line with Nicod’s
theory. It says that, if evidence e confirms “all Fs are G s”, then it also mention of “the absence of G [here, non-black as evidence] in a case of F
confirms that a further object will be G , if it is F. Notably, this does not [here, raven as an auxiliary assumption]”. It is also true that an object that
hold for HD-confirmation. Here is why. Given k = Fa (i.e., the is found not to be a raven HD-confirms h, but only relative to
assumption that a comes from the F population), we have that e = Ga k = ¬black(a) , that is, if a is assumed to have been taken among non-
HD-confirms h = ∀x(Fx → Gx), because the latter entails the former black objects to begin with; and this seems acceptable too (after all, while
(given k ). (That’s the HD reconstruction of Nicod’s insight, see below.) sampling from non-black objects, one might have found the
We also have, of course, that h entails h∗ = Fb → Gb. And yet, contrary counterinstance of a raven, but didn’t). Unlike Hempel’s theory, moreover,
to (PIC), since h∗ does not entail e (given k), it is not HD-confirmed by it HD-confirmation does not yield the debatable implication that, by itself
either. The troubling conclusion is that the observation that a swan is white (that is, given k = ⊤ ), the observation of a non-raven a , ¬raven(a), must
(or that a million of them are, for that matters) does not HD-confirm the confirm h .
prediction that a further swan will be found to be white.
Interestingly, the introduction of auxiliary hypotheses and assumptions
2.2 Back to black (ravens) shows that the issues surrounding Nicod’s remarks can become
surprisingly subtle. Consider the following statements (Maher’s 2006
One attractive feature of HD-confirmation is that it largely eludes the example):
ravens paradox. As the hypothesis h that all ravens are black does not
entail that some generally sampled object a will be a black raven, the HD (α1 ) ∀x(white(x) → ¬black(x))
view of confirmation is not committed to the eminently Hempelian (α2 ) ∃x(swan(x)) → ∃y(swan(y) ∧ black(y))
implication that e = raven(a) ∧ black(a) confirms h. Likewise,
α1 simply specifies that no object is both white and black, while α2 says
¬black(a) ∧ ¬raven(a) does not HD-confirm that all non-black objects
that, if there are swans at all, then there also is some black swan. Also
are non-raven. The derivation of the paradox, as presented above, is thus
posit, again, e = swan(a) ∧ white(a) . Under α1 and α2 , the observation of
blocked.
a white swan clearly disconfirms (indeed, refutes) the hypothesis h that all
Indeed, HD-confirmation yields a substantially different reading of swans are white. Hempel’s theory (extended) faces difficulties here,
Nicod’s insight as compared to Hempel’s theory (Okasha 2011 has an because for k = deve (α1 ∧ α2 ) it turns out that e ∧ k is inconsistent. But
important discussion of this distinction). Here is how it goes. If object a is HD-confirmation gets this case right, thus capturing appropriate boundary
assumed to have been taken among ravens—so that, crucially, the conditions for Nicod’s generally sensible claims. For, with k = α1 ∧ α2 ,

one has that h ∧ k is consistent and entails ¬e (for it entails that no swan Finally, it should be noted that HD-confirmation offers no substantial relief
exists), so that e HD-disconfirms (refutes) h relative to k (see Good 1967 from the blite paradox. On the one hand,
for another famous counterexample to Nicod’s condition). e = raven(a) ∧ ext≤T (a) ∧ black(a) does not, as such, HD-confirm either
h = ∀x(raven(x) → black(x)) or h∗ = ∀x(raven(x) → blite(x)), that is,
HD-confirmation, however, is also known to suffer from distinctive for empty k . On the other hand, if object a is assumed to have been
“paradoxical” implications. One of the most frustrating is surely the sampled from ravens before T (that is, given k = raven(a) ∧ ext≤T (a)),
following (see Osherson, Smith, and Shafir 1986, 206, for further specific then black(a) is entailed by both “all ravens are black” and “all ravens are
problems). blite” and therefore HD-confirms each of them. So HD-confirmation, too,
sanctions the existence of confirmation relations that seem intuitively
The irrelevant conjunction paradox. Suppose that e confirms h relative to
(possibly empty) k . Let statement q be logically consistent with e ∧ h ∧ k,
unsound (indeed, indefinitely many of them: as we know, other variations
of h∗ can be conceived at will, like the “blurple” hypothesis). One could
but otherwise entirely irrelevant for all of those conjuncts (perhaps
insist that HD does handle the blite paradox after all, because black(a)
belonging to a completely separate domain of inquiry). Does e confirm
(given k as above) does not HD-confirms that a raven will be white if
h ∧ q (relative to k ) as it does with h? One would want to say no, and this
examined after T (Kuipers 2000, 29 ff.). Unfortunately (as pointed out by
implication can be suitably reconstructed in Hempel’s theory. HD-
Schurz 2005, 148) black(a) does not HD-confirm that a raven will be
confirmation, on the contrary, can not draw this distinction: it is easy to
black if examined after T either (again, given k as above). That’s because,
show that, on the conditions specified, if the HD clause for confirmation is
satisfied for e and h (given k), so it is for e and h ∧ q (given k ). (This is
as already pointed out, HD-confirmation fails the predictive inference
simply because, if h ∧ k ⊨ e, then h ∧ q ∧ k ⊨ e , too, by the monotonicity

condition (PIC) in general. So, all in all, HD-confirmation can not tell
black from blite any more than Hempel-confirmation can.
of classical logical entailment.)
Kuipers (2000, 25) suggested that one can live with the irrelevant 2.3 Underdetermination and the Duhemian challenge
conjunction problem because, on the conditions specified, e would still not
HD-confirm q alone (given k ), so that HD-confirmation can be The issues above look contrived and artificial to some people’s taste—
“localized”: h is the only bit of the conjunction h ∧ q that gets any even among philosophers. Many have suggested a closer look at real-
confirmation on its own, as it were. Other authors have been reluctant to world inferential practices in the sciences as a more appropriate
bite the bullet and have engaged in technical refinements of the “naïve” benchmark for assessment. For one thing, the very idea of hypothetico-
HD view. In these proposals, the spread of HD-confirmation upon deductivism has often been said to stem from the origins of Western
frivolous conjunctions can be blocked at the expense of some additional science. As reported by Simplicius of Cilicia (sixth century A.D.) in his
logical machinery (see Gemes 1993, 1998; Schurz 1991, 1994). commentary on Aristotle’s De Caelo, Plato had challenged his pupils to
identify combinations of “ordered” motions by which one could account
for (namely, deduce) the planets’ wandering trajectories across the heavens

as observed by the Earth. As a matter of historical fact, mathematical are logically consistent with any piece of checkable evidence. Unless, of
astronomy has engaged in just this task for centuries: scholars have been course, the logical connection is underpinned by auxiliary hypotheses and
trying to define geometrical models from which the apparent motion of assumptions suitably bridging the gap between the observational and non-
celestial bodies would derive. observational vocabulary, as it were. But then, once auxiliaries are in play,
logic alone guarantees that some k exists such that h ∧ k is consistent,
It is fair to say that, at its roots, the kind of challenges that the HD h ∧ k ⊨ e, and k ⊭ e , so that confirmation holds in naïve HD terms (that’s
framework faces with scientific reasoning is not so different from the main just the UT result above). Apparently, when Duhem’s point applies, the
puzzles that arise from philosophical considerations of a more formal uncritical supporter of whatever hypothesis h can legitimately claim
kind. Still, the two areas turn out to be complementary in important ways. (naïve HD) confirmation from any e by simply shaping k conveniently. In
The following statement will serve as a useful starting point to extend the this sense, hypothesis assessment would be radically “underdetermined”
scope of our discussion. by any amount of evidence practically available.
Underdetermination Theorem (UT) for “naïve” HD-confirmation Influential authors such as Thomas Kuhn (1962/1970) (but see Laudan
For any contingent h, e ∈ L , if h and e are logically consistent, there 1990, 268, for a more extensive survey) relied on Duhemian insights to
exists some k ∈ L such that e HD-confirms h relative to k . suggest that confirmation by empirical evidence is too weak a force to
drive the evaluation of theories in science, often inviting conclusions of a
(UT) is an elementary logical fact that has been long recognized (see, e.g.,
Glymour 1980a, 36). In purely formal terms, just positing k = h → e will
relativistic flavor (see Worrall 1996 for an illuminating reconstruction
along these lines). Let us briefly consider a classical case, which Duhem
do for a proof. To appreciate how (UT) can spark any philosophical
himself thoroughly analyzed: the wave vs. particle theories of light in
interest, one has to combine it with some insightful remarks first put
modern optics. Across the decades, wave theorists were able to deduce an
forward by Pierre Duhem (1906) and then famously revived by Quine
impressive list of important empirical facts from their main hypothesis
(1951) in a more radical style. (Indeed, (UT) essentially amounts to the
along with appropriate auxiliaries, diffraction phenomena being only one
“entailment version” of “Quinean underdetermination” in Laudan 1990,
major example. But many particle theorists’ reaction was to retain their
274.)
hypothesis nonetheless and to reshape other parts of the “theoretical maze”
Duhem (he himself a supporter of the HD view) pointed out that in mature (i.e., k ; the term is Popper’s, 1963, p. 330) to recover those observed facts
sciences such as physics most hypotheses or theories of real interest can as consequences of their own proposal. And as we’ve seen, if the bare
not be contradicted by any statement describing observable states of logic of naïve HD was to be taken strictly, surely they could have claimed
affairs. Taken in isolation, they simply do not logically imply, nor rule out, their overall hypothesis to be confirmed too, just as much as their
any observable fact, essentially because (unlike “all ravens are black”) opponents.
they involve the mention of unobservable entities and processes. So, in
Importantly, they didn’t. In fact, it was quite clear that particle theorists,
effect, Duhem emphasized that, typically, scientific hypotheses or theories
unlike their wave-theory opponents, were striving to remedy weaknesses

rather than scoring successes (see Worrall 1990). But why, then? Because, for Newtonian physics, all the way up to the Higgs boson for so-called
as Duhem himself clearly realized, the logic of naïve HD “is not the only standard model of subatomic particles.
rule for our judgments” (1906, 217). The lesson of (UT) and the
Duhemian insight is not quite, it seems, that naïve HD is the last word and The predictivist solution to the underdetermination problem is fairly
scientific inference is unconstrained by stringent rational principles, but radical: many of the relevant factual consequences of h ∧ k will be already
rather that the HD view has to be strengthened in order to capture the real known when this theory is articulated, and so unfit for confirmation.
nature of evidential support in rational scientific inference. At least, that’s Critics have objected that predictivism is in fact far too restrictive. There
the position of a good deal of philosophers of science working within the seem to be many cases in which already known phenomena clearly do
HD framework broadly construed. It has even been maintained that “no provide support to a new hypothesis or theory. Zahar (1973) first raised
serious twentieth-century methodologist” has ever subscribed to the naïve this problem of “old evidence”, then made famous by Glymour (1980a, 85
HD view above “without crucial qualifications” (Laudan 1990, 278; also ff.) as a difficulty for Bayesianism (see Section 3 below). Examples of this
see Laudan and Leplin 1991, 466). kind abound in the history of science as elsewhere, but the textbook
illustration has become the precession of Mercury’s perihelion, a lasting
So the HD approach to confirmation has yielded a number of more anomaly for Newtonian physics: Einstein’s general relativity calculations
articulated variants to meet the challenge of underdetermination. got this long-known fact right, thereby gaining a remarkable piece of
Following (loosely) Norton (2005), we will now survey an instructive initial support for the new theory. In addition to this problem with old
sample of them. evidence, HD predictivism also seems to lack a principled rationale. After
all, the temporal order of the discovery of e and of the articulation of h
2.4 The extended HD menu and k may well be an entirely accidental historical contingency. Why
should it bear on the confirmation relationship among them? (See Giere
Naïve HD can be enriched by a resolute form of predictivism. According 1983 and Musgrave 1974 for classical discussions of these issues. Douglas
to this approach, the naïve HD clause for confirmation is too weak because and Magnus 2013 and Barnes 2018 offer more recent views and rich lists
e must have been predicted in advance from h ∧ k. Karl Popper’s of further references.)
(1934/1959) account of the “corroboration” of hypotheses famously
embedded this view, but squarely predictivist stances can be traced back to As a possible response to the difficulties above, naïve HD can be enriched
early modern thinkers like Christiaan Huygens (1629–1695) and Gottfried by the use-novelty criterion (UN) instead. The UN reaction to the
Wilhelm Leibniz (1646–1716), and in Duhem’s work itself. The underdetermination problem is more conservative than the temporal
predictivist sets a high bar for confirmation. Her favorite examples predictivist strategy. According to this view, to improve on the weak naïve
typically include stunning episodes in which the existence of previously HD clause for confirmation one only has to rule out one particular class of
unknown objects, phenomena, or whole classes of them is anticipated: the cases, i.e., those in which the description of a known fact, e, served as a
phases of Venus for Copernican astronomy or the discovery of Neptune constraint in the construction of h ∧ k. The UN view thus comes equipped
with a rationale. If h ∧ k was shaped on the basis of e , UN advocates point

out, then it was bound to get that state of affairs right; the theory never ran safely ruled out. The founding figure of eliminativism is Francis Bacon
any risk of failure, thus did not achieve any particularly significant success (1561–1626). John Stuart Mill (1843/1872) is a major representative in
either. Precisely in these cases, and just for this reason, the evidence e later times, and Deborah Mayo’s “error-statistical” approach to hypothesis
must not be double-counted: by using it for the construction of the theory, testing arguably develops this tradition (Mayo 1996 and Mayo and Spanos
its confirmational power becomes “dried out”, so to speak. 2010; see Bird 2010, Kitcher 1993, 219 ff., and Meehl 1990 for other
contemporary variations).
The UN completion of naïve HD originated from Lakatos and some of his
collaborators (see Lakatos and Zahar 1975 and Worrall 1978; also see Eliminativism is most credible when experimentation is at issue (see, e.g.,
Giere 1979, 161–162, and Gillies 1989 for similar views), although Guala 2012). Indeed, the appeal to Bacon’s idea of crucial experiment
important hints in the same direction can be found at least in the work of (instantia crucis) and related notions (e.g., “severe testing”) is a fairly
William Whewell (1840/1847). Consider the touchstone example of reliable mark of eliminativist inclinations. Experimentation is, to a large
Mercury again. According to Zahar (1973), Einstein did not need to rely extent, precisely an array of techniques to keep undesired interfering
on the Mercury data to define theory and auxiliaries as to match factors at a minimum by active manipulation and deliberate control (think
observationally correct values for the perihelion precession (also see of the blinding procedure in medical trials, with h the hypothesized
Norton 2011a; and Earman and Janssen 1993 for a very detailed, and more effectiveness of a novel treatment and e a relative improvement in clinical
nuanced, account). Being already known, the fact was not of course endpoints for a target subsample of patients thus treated). When this kind
predicted in a strictly temporal sense, and yet, on Zahar’s reading, it could of control obtains, popular statistical tools are supposed to allow for the
have been: it was “use-novel” and thus fresh for use to confirm the theory. calculation of the probability of e in case h is false meant as a “relative
For a more mundane illustration, so-called cross-validation techniques frequency in a (real or hypothetical) series of test applications” (Mayo
represent a routine application of the UN idea in statistical settings (as 1991, 529), and to secure a sufficiently low value to validate the positive
pointed out by Schurz 2014, 92; also see Forster 2007, 592 ff.). According outcome of the test. It is much less clear how firm a grip this approach can
to some commentators, however, the UN criterion needs further retain when inference takes place at higher levels of generality and
elaboration (see Hitchcock and Sober 2004 and Lipton 2005), while others theoretical commitment, where the hypothesis space is typically much too
have criticized it as essentially wrong-headed (see Howson 1990 and poorly ordered to fit routine error-statistical analyses. Indeed, Laudan
Mayo 1991, 2014; also see Votsis 2014). (1997, 315; also see Musgrave 2010) spotted in this approach the risk of a
“balkanization” of scientific reasoning, namely, a restricted focus on
Yet another way to enrich naïve HD is to combine it with eliminativism. scattered pieces of experimental inference (but see Mayo 2010 for a
According to this view, the naïve HD clause for confirmation is too weak defense).
because there must have been a low (enough) objective chance of getting
the outcome e (favorable to h) if h was false, so that few possibilities exist Naïve HD can also be enriched by the notion of simplicity. According to
that e may have occurred for some reason other than the truth of h . Briefly this view, the naïve HD clause for confirmation is too weak because h ∧ k
put, the occurrence of e must be such that most alternatives to h can be must be a simple (enough), unified way to account for evidence e . A

classic reference for the simplicity view is Newton’s first law of scale theories and relatively high-level kinds of evidence. Dealing with
philosophizing in the Principia (“admit no more causes of natural things Einstein’s general relativity, for instance, Janssen (2003) greatly
than such as are both true and sufficient to explain their appearances”), emphasizes its explanation of the equivalence of inertial and gravitational
echoing very closely Ockham’s razor. This basic idea has never lost its mass (essentially a brute fact in Newtonian physics) over the resolution of
appeal—even up to recent times (see, e.g., Quine and Ullian 1970, 69 ff.; the puzzle of Mercury’s perihelion. Explanationist accounts are also
Sober 1975; Zellner, Keuzenkamp, and McAleer 2002; Scorzato 2013). distinctively well-equipped to address inference patterns from non-
experimental sciences (Cleland 2011).
Despite Thomas Kuhn’s (1957, 181) suggestions to the contrary, the
success of Copernican astronomy over Ptolemy’s system has remained an The problems faced by these approaches are similar to those affecting the
influential case study fostering the simplicity view (Martens 2009). simplicity view. Agreement is still lacking on the nature of scientific
Moreover, in ordinary scientific problems such as curve fitting, formal explanation (see Woodward 2019) and it is not clear how far an
criteria of model selection are applied where the paucity of parameters can explanationist variant of HD can go without a sound analysis of that
be interpreted naturally as a key dimension of simplicity (Forster and notion. Moreover, some critics have wondered why the relationship of
Sober 1994). Traditionally, two main problems have proven pressing, and confirmation should be affected by an explanatory connection with the
frustrating, for the simplicity approach. First, how to provide a sufficiently evidence per se (see Salmon 2001).
coherent and illuminating explication of this multifaceted and elusive
notion (see Riesch 2010); and second, how to justify the role of simplicity The above discussion does not display an exhaustive list (nor are the listed
as a properly epistemic (rather than merely pragmatic) virtue (see Kelly options mutually exclusive, for that matter: see, e.g., Baker 2003; also see
2007, 2008). Worrall 2010 for some overlapping implications in an applied setting of
real practical value). And our sketched presentation hardly allows for any
Finally, naïve HD can be enriched by the appeal to explanation. Here, the conclusive assessment. It does suggest, however, that reports of the death
naïve HD clause for confirmation is meant to be too weak because h ∧ k of hypothetico-deductivism (see Earman 1992, 64, and Glymour 1980b)
must be able (not only to entail, but) to explain e . By this move, the HD might have been exaggerated. For all its difficulties, HD has proven fairly
approach embeds the slogan of the so-called inference to the best resilient at least as a basic framework to elucidate some key aspects of
explanation view: “observations support the hypothesis precisely because how hypotheses can be confirmed by the evidence (see Betz 2013, Gemes
it would explain them” (Lipton 2000, 185; also see Lipton 2004). 2005, and Sprenger 2011b for consonant points of view).
Historically, the main source for this connection between explanation and
support is found in the work of Charles Sanders Peirce (1839–1914). 3. Bayesian confirmation theories
Janssen (2003) offers a particularly neat contemporary exhibit, explicitly
aimed at “curing cases of the Duhem-Quine disease” (484; also see Bayes’s theorem is a very central element of the probability calculus (see
Thagard 1978, and Douven 2017 for a relevant survey). Quite unlike Joyce 2019). For historical reasons, Bayesian has become a standard label
eliminativist approaches, explanationist analyses tend to focus on large- to allude to a range of approaches and positions sharing the common idea

that probability (in its modern, mathematical sense) plays a crucial role in approaches, depending on the circumstances, P may (or should) be set up
rational belief, inference, and behavior. According to Bayesian so that portions of the evidence available are in fact bracketed.
epistemologists and philosophers of science, (i) rational agents have (Unsurprisingly, further subtleties arise as soon as one delves a bit further
credences differing in strength, which moreover (ii) satisfy the probability into the precise meaning and scope of TE; see Fitelson 2008 and
axioms, and can thus be represented in probabilistic form. (In non- Williamson 2002, Chs. 9–10, for important discussions.)
Bayesian models (ii) is rejected, but (i) may well be retained: see Huber
and Schmidt-Petri 2009, Levi 2008, and Spohn 2012.) Well-known Of course, many intermediate positions exist between extreme forms of
arguments exist in favor of this position (see, e.g., Easwaran 2011a; permissivism and impermissivism so outlined, and more or less the same
Pettigrew 2016; Skyrms 1987; Vineberg 2016), although there is no lack applies for the TE issue. The above distinctions are surely rough enough,
of difficulties and criticism (see, e.g., Easwaran 2011b; Hájek 2008; Kelly but useful nonetheless. Impermissive TE Bayesianism has served as a
and Glymour 2004; Norton 2011b). received view in early Bayesian philosophy of science (e.g., Carnap
1950/1962). But impermissivism is easily found in combination with non-
Beyond the core ideas above, however, the theoretical landscape of TE positions, too (see, e.g., Maher 1996). TE permissivism seems a good
Bayesianism is quite as hopelessly diverse as it is fertile. Surveys and state approximation of De Finetti’s (2008) stance, while non-TE permissivism
of art presentations are already numerous, and ostensibly growing (see, is arguably close to a standard view nowadays (see, e.g., Howson and
e.g., Good 1971; Joyce 2011; Oaksford and Chater 2007; Sprenger and Urbach 2006). No more than this will be needed to begin our exploration
Hartmann 2020; Weisberg 2015). For the present purposes, attention can of Bayesian confirmation theories.
be restricted to a classification that is still fairly coarse-grained, and based
on just two dimensions or criteria. 3.1 Probabilistic confirmation as firmness
First, there is a distinction between permissivism and impermissivism (see Let us posit a set P of probability functions representing possible states of
Meacham 2014 and Kopec and Titelbaum 2016 for this terminology). For belief about a domain that is described in a finite language L with L the
permissive Bayesians (often otherwise labelled “subjectivists”), set of its closed sentences. From now on, unless otherwise specified,
accordance with the probability axioms is the only clear-cut constraint on whenever considering some h, e, k ∈ L and P ∈ P, we will invariably rely
the credences of a rational agent. In impermissive forms of Bayesianism on the following provisos:
(often otherwise called “objective”), further constraints are put forward
that significantly restrict the range of rational credences, possibly up to i. both e ∧ k and h ∧ k are consistent;
one single “right” probability function in any given setting. Second, there ii. P(e ∧ k), P(h ∧ k) > 0;
are different attitudes towards so-called principle of total evidence (TE) iii. P(k) > P(h ∧ k) (unless k ⊨ h);
for the credences on which a reasoner relies. TE Bayesians maintain that iv. P(e ∧ k) > P(e ∧ h ∧ k) (unless e ∧ k ⊨ h); and
the relevant credences should be represented by a probability function P v. P(e ∧ h ∧ k) > 0, as long as e ∧ h ∧ k is consistent.
which conveys the totality of what is known to the agent. For non-TE

(These assumptions are convenient and critical for technical reasons, but (P1) Final probability
not entirely innocent. Festa 1999 and Kuipers 2000, 44 ff., discuss some For any h, e1 , e2 , k ∈ L and any P ∈ P, CP (h, e1 ∣ k) ⋛ CP (h, e2 ∣ k)
limiting cases that are left aside here owing to these constraints.) if and only if P(h ∣ e1 ∧ k) ⋛ P(h ∣ e2 ∧ k).
A probabilistic theory of confirmation can be spelled out through the (P1) is itself a comparative, or ordinal, principle, stating that, for any fixed
definition of a function CP (h, e ∣ k) : {L3 × P} → ℜ representing the hypothesis h, the final (or posterior) probability and confirmation always
degree of confirmation that hypothesis h receives from evidence e relative move in the same direction in the light of data, e (given k ). Interestingly,
to k and probability function P. CP (h, e ∣ k) will then have relevant (P0) and (P1) are already sufficient to single out one traditional class of
probabilities as its building blocks, according to the following basic measures of probabilistic confirmation, if conjoined with the following
postulate of probabilistic confirmation: (see Crupi and Tentori 2016, 656, Schippers 2017, and also Törnebohm
1966, 81):
(P0) Formality
There exists a function g such that, for any h, e, k ∈ L and any P ∈ P, (P2) Local equivalence
CP (h, e ∣ k) = g[P(h ∧ e ∣ k), P(h ∣ k), P(e ∣ k)] . For any h1 , h2 , e, k ∈ L and any P ∈ P, if h1 and h2 are logically
equivalent given e and k, then CP (h1 , e ∣ k) = CP (h2 , e ∣ k).
Note that the probability distribution over the algebra generated by h and
e , conditional on k , is entirely determined by P(h ∧ e ∣ k) , P(h ∣ k) and The following can then be shown:
P(e ∣ k) . Hence, (P0) simply states that CP (h, e ∣ k) depends on that
distribution, and nothing else. (The label for this assumption is taken from Theorem 1
Tentori, Crupi, and Osherson 2007, 2010.) (P0), (P1) and (P2) hold if and only if there exists a strictly increasing
function f such that, for any h, e, k ∈ L and any P ∈ P,
Hempelian and HD confirmation, as discussed above, are qualitative CP (h, e ∣ k) = f [P(h ∣ e ∧ k)].
theories of confirmation. They only tell us whether evidence e confirms
(disconfirms) hypothesis h given k. However, assessments of the amount Theorem 1 provides a simple axiomatic characterization of the class of
of support that some evidence brings to a hypothesis are commonly confirmation functions that are strictly increasing with the final probability
involved in scientific reasoning, as well as in other domains, if only in the of the hypothesis given the evidence (and k ) (proven in Schippers 2017).
form of comparative judgments such as “hypothesis h is more strongly All the functions in this class are ordinally equivalent, meaning that they
confirmed by e1 than by e2 ” or “e confirms h1 to a greater extent than h2 ”. imply the same rank order of CP (h, e ∣ k) and CP∗ (h∗ , e∗ ∣ k ∗ ) for any
Consider, for instance, the following principle, a veritable cornerstone of h, h∗ , e, e∗ , k, k ∗ ∈ L and any P, P∗ ∈ P.
By (P0), (P1) and (P2), we thus have CP (h, e ∣ k) = f [P(h ∣ e ∧ k)],

probabilistic confirmation in all of its variations (see Crupi, Chater, and
Tentori 2013 for a list of references):
implying that the more likely h is given the evidence the more it is
confirmed. This approach explicates confirmation precisely as the overall

credibility of a hypothesis (firmness is Carnap’s 1950/1962 telling term, (false). (Sometimes a threshold higher than a probability 1⁄2 is identified,
xvi). In this view, “Bayesian confirmation theory is little more than the but this complication would add little for our present purposes.)
examination of [the] properties” of the posterior probability function
(Howson 2000, 179). The ordinal notion of confirmation is of high theoretical significance
because ordinal divergences, unlike purely quantitative differences, imply
As we will see, the ordinal level of analysis is a solid and convenient opposite comparative judgments for some evidence-hypothesis pairs. A
middleground between a purely qualitative and a thoroughly quantitative refinement from the ordinal to a properly quantitative level is also be of
(metric) notion of confirmation. To begin with, ordinal notions are in interest, however, and much useful for tractability and applications. For
general sufficient to move “upwards” to the qualitative level as follows: example, one can have 0 as a convenient neutrality threshold for
confirmation as firmness, provided that the following functional
Qualitative confirmation from ordinal relations (QC) representation is adopted (see Peirce 1878 for an early occurrence):
For any h, e, k ∈ L and any P ∈ P:
P(h ∣ e ∧ k)
F(h, e ∣ k) = log[
e CP -confirms h relative to k if and only if P(¬h ∣ e ∧ k) ]
CP (h, e ∣ k) > CP (¬h, e ∣ k); = log Odds(h ∣ e ∧ k)
e CP -disconfirms h relative to k if and only if
CP (h, e ∣ k) < CP (¬h, e ∣ k); (The base of the logarithm can be chosen at convenience, as long as it is
e is CP -neutral for h relative to k if and only if strictly greater than 1.)
CP (h, e ∣ k) = CP (¬h, e ∣ k).
A quantitative requirement that is often put forward is the following
Given Theorem 1, (P0), (P1) and (P2) can be combined with the stringent form of additivity:
definitions in (QC) to derive the following qualitative notion of
Strict additivity (SA)
For any h, e1 , e2 , k ∈ L and any P ∈ P,
probabilistic confirmation as firmness:
Confirmation as firmness (F-confirmation, qualitative) CP (h, e1 ∧ e2 ∣ k) = CP (h, e1 ∣ k) + CP (h, e2 ∣ e1 ∧ k).

For any h, e, k ∈ L and any P ∈ P:
Although extraneous to F-confirmation, Strict Additivity will prove of use
e F-confirms h relative to k if and only if P(h ∣ e ∧ k) > 1⁄2 ; later on for the discussion of further variants of Bayesian confirmation
e F-disconfirms h relative to k if and only if P(h ∣ e ∧ k) < 1⁄2 ; theory.
e is F-neutral for h relative to k if and only if P(h ∣ e ∧ k) = 1⁄2 .
The point of qualitative F-confirmation is thus straightforward: h is said to

be (dis)confirmed by e (given k) if it is more likely than not to be true
1
2

3.2 Strengths and infirmities of firmness F(h, e ∣ k) instantiates Confirmation Complementarity in a simple and
elegant way, that is, it satisfies CP (h, e ∣ k) = −CP (¬h, e ∣ k).
Confirmation as firmness shares a number of structural properties with
Hempelian confirmation. It satisfies the Special Consequence Condition, F-confirmation also implies another attractive quantitative result,
thus the Predictive Inference Condition too. It satisfies the Entailment alleviating the ailments of the irrelevant conjunction paradox. In the
Condition and, in virtue of (P1), extends it smoothly to the following statement below, indicating this result, the irrelevance of q for hypothesis
ordinal counterpart: h and evidence e (relative to k ) is meant to amount to the probabilistic
independence of q from h, e and their conjunction (given k), that is, to
Entailment condition (ordinal extension) (EC-Ord) P(h ∧ q ∣ k) = P(h ∣ k)P(q ∣ k), P(e ∧ q ∣ k) = P(e ∣ k)P(q ∣ k) , and
For any h, e1 , e2 , k ∈ L and any P ∈ P such that k ⊭ h : P(h ∧ e ∧ q ∣ k) = P(h ∧ e ∣ k)P(q ∣ k) , respectively.
i. if, e1 ∧ k ⊨ h and e2 ∧ k ⊭ h , then h is more confirmed by e1 Confirmation upon irrelevant conjunction (ordinal solution) (CIC)
than by e2 relative to k , that is, CP (h, e1 ∣ k) > CP (h, e2 ∣ k); For any h, e, q, k ∈ L and any P ∈ P, if e confirms h relative to k and
ii. if, e1 ∧ k ⊨ h and e2 ∧ k ⊨ h, then h is equally confirmed by e1 q is irrelevant for h and e relative to k, then
and by e2 relative to k , that is, CP (h, e1 ∣ k) = CP (h, e2 ∣ k). CP (h, e ∣ k) > CP (h ∧ q, e ∣ k).
According to (EC-Ord) not only is classical entailment retained as a case So, even in case it is qualitatively preserved across the tacking of q onto h,
of confirmation, it also represents a limiting case: it is the strongest the positive confirmation afforded by e is at least bound to quantitatively
possible form of confirmation that a fixed hypothesis h can receive. decrease thereby.
F-confirmation also satisfies Confirmation Complementarity and, Partly because of appealing formal features such as those mentioned so
moreover, extends it to its appealing ordinal counterpart (see Crupi, Festa, far, there is a long list of distinguished scholars advocating the firmness
and Buttasi 2010, 85–86), that is: view of confirmation, from Keynes (1921) and Hosiasson-Lindenbaum
(1940) onwards, most often coupled with some form of impermissive
Confirmation complementarity (ordinal extension) (CC-Ord) Bayesianism (see Hawthorne 2011 and Williamson 2011 for contemporary
CP (¬h, e ∣ k) is a strictly decreasing function of CP (h, e ∣ k), that is, variations). In fact, F-confirmation fits most neatly a classical form of TE
for any h, h∗ , e, e∗ , k ∈ L and any P ∈ P, CP (h, e ∣ k) ⋛ CP (h∗ , e∗ ∣ k) impermissivism à la Carnap, where one assumes that k = ⊤, that P is an
if and only if CP (¬h, e ∣ k) ⋚ CP (¬h∗ , e∗ ∣ k). “objective” initial probability based on essentially logical considerations,
and that all the non-logical information available is collected in e . The
(CC-Ord) neatly reflects Keynes’ (1921, 80) remark that “an argument is
spirit of the Carnapian project never lost its appeal entirely (see, e.g., Festa
always as near to proving or disproving a proposition, as it is to disproving
2003, Franklin 2001, Maher 2010, Paris 2011). However, the idea of a
or proving its contradictory”. Indeed, quantitatively, the measure
“logical” interpretation of P got stuck into difficulties that are often seen

as insurmountable (e.g., Earman and Salmon 1992, 85–89; Gillies 2000, any of a number of alternative viral diagnoses. This judgment clashes with
Ch. 3; Hájek 2019; Howson and Urbach 2006, 59–72; van Fraassen 1989, (Cons), though, which then seems an overly strong constraint.
Ch. 12; Zabell 2011). And arguably, lacking some robust and effective
impermissivist policy, the account of confirmation as firmness ends up Notably, (Cons) was defended by Hempel (1945) and, in fact, one can
loosing much of its philosophical momentum. The issues surrounding the show that it follows from the conjunction of (qualitative) Confirmation
ravens and blite paradoxes provide a useful illustration. Complementary and the Special Consequence Condition, and so from both
Hempelian and F-confirmation. This is but one sign of how stringent the
Consider again h = ∀x(raven(x) → black(x)), and the main analyses of Special Consequence Condition is. Mainly because of the latter, both the
“the observation that a is a black raven” encountered so far, that is: Hempelian and the firmness views of confirmation must depart from the
plausible HD idea that hypotheses are generally confirmed by their
i. k = ⊤ and e = raven(a) ∧ black(a), and verified consequences (see Hempel 1945, 103–104). We will come back to
ii. k = raven(a) and e = black(a). this while discussing our next topic: a very different Bayesian explication
of confirmation, based on the notion of probabilistic relevance.
In both cases, whether e F-confirms h or not (relative to k) critically
depends on P: if the prior P(h ∣ k) is low enough, e won’t do no matter
what under either (i) or (ii); and if it is high enough, h will be F -confirmed
3.3 Probabilistic relevance confirmation
either way. As a consequence, the F-confirmation view, by itself, does not
We’ve seen that the firmness notion of probabilistic confirmation can be
offer any definite hint as to when, how, and why Nicod’s remarks apply or
singled out through one ordinal constraint, (P2), in addition to the
not.
fundamental principles (P0)–(P1). The counterpart condition for the so-
For the purposes of our discussion, the following condition reveals another called relevance notion of probabilistic confirmation is the following:
debatable aspect of the firmness explication of confirmation.
(P3) Tautological evidence
Consistency condition (Cons) For any h1 , h2 , k ∈ L and any P ∈ P, CP (h1 , ⊤ ∣ k) = CP (h2 , ⊤ ∣ k).
For any h, h∗ , e, k ∈ L and any P ∈ P, if k ⊨ ¬(h ∧ h∗ ) then e
(P3) implies that any hypothesis is equally “confirmed” by empty
confirms h given k if and only if e disconfirms h∗ given k .
evidence. We will say that CP (h, e ∣ k) represents the probabilistic
(Cons) says that evidence e can never confirm incompatible hypotheses. relevance notion of confirmation, or relevance-confirmation, if and only if
But consider, by way of illustration, a clinical case of an infectious disease it satisfies (P0), (P1) and (P3). These conditions are sufficient to derive the
of unknown origin, and suppose that e is the failure of antibiotic treatment. following, purely qualitative principle, according to the definitional
Arguably, there is nothing wrong in saying that, by discrediting bacteria as method in (QC) above (see Crupi and Tentori 2014, 82, and Crupi 2015).
possible causes, the evidence confirms (viz. provides some support for)
Probabilistic relevance confirmation (qualitative)
For any h, e, k ∈ L and any P ∈ P :

e relevance-confirms h relative to k if and only if The qualitative notion of relevance confirmation already has some
P(h ∣ e ∧ k) > P(h ∣ k); interesting consequences. It implies, for instance, the following
e relevance-disconfirms h relative to k if and only if remarkable fact:
P(h ∣ e ∧ k) < P(h ∣ k);
e is relevance-neutral for h relative to k if and only if Complementary Evidence (CompE)
P(h ∣ e ∧ k) = P(h ∣ k). For any h, e, k ∈ L and any P ∈ P, e confirms h relative to k if and
only if ¬e disconfirms h relative to k.
The point of relevance confirmation is that the credibility of a hypothesis
can be changed in either a positive (confirmation in a strict sense) or The importance of (CompE) can be illustrated as follows. Consider the
negative way (disconfirmation) by the evidence concerned (given k). case of a father suspected of abusing his son. Suppose that the child does
Confirmation (in the strict sense) thus reflects an increase from initial to claim that s/he has been abused (label this evidence e ). A forensic
final probability, whereas disconfirmation reflects a decrease (see psychiatrist, when consulted, declares that this confirms guilt (h) .
Achinstein 2005 for some diverging views on this very idea). Alternatively, suppose that the child is asked and does not report having
been abused (¬e). As pointed out by Dawes (2001), it may well happen
The qualitative notions of confirmation as firmness and as relevance are that a forensic psychiatrist will nonetheless interpret this as evidence
demonstrably distinct. Unlike firmness, relevance confirmation can not be confirming guilt (suggesting that violence has prompted the child’s
formalized by the final probability alone, or any increasing function denial). One might want to argue that, other things being equal, this kind
thereof. To illustrate, the probability of an otherwise very rare disease (h) of “heads I win, tails you lose” judgment would be inconsistent, and thus
can be quite low even after a relevant positive test result (e); yet h is in principle untenable. Whoever concurs with this line of argument (as
relevance-confirmed by e to the extent that its probability rises thereby. By Dawes 2001 himself did) is likely to be relying on the relevance notion of
the same token, the probability of the absence of the disease (¬h) can be confirmation. In fact, no other notion of confirmation considered so far
quite high despite the positive test result (e) , yet ¬h is relevance- provides a general foundation for this judgment. F-confirmation, in
disconfirmed by e to the extent that its probability decreases thereby. particular, would not do, for it does allow that both e and ¬e confirm h
Perhaps surprisingly, the distinction between firmness and relevance (relative to k ). This is because, mathematically, it is perfectly possible for
confirmation—“extremely fundamental” and yet “sometimes unnoticed”, both P(h ∣ e ∧ k) and P(h ∣ ¬e ∧ k) to be arbitrarily high above 1⁄2 .
as Salmon (1969, 48–49) put it—had to be stressed time and again to Condition (CompE), on the contrary, ensures that only one between the
achieve theoretical clarity in philosophy (e.g., Popper 1954; Peijnenburg complementary statements e and ¬e can confirm hypothesis h (relative to
2012) as well as in other domains concerned, such as artificial intelligence k ). (To be precise, HD-confirmation also satisfies condition CompE, yet it
and the psychology of reasoning (see Horvitz and Heckerman 1986; would fail the above example all the same, although for a different reason,
Crupi, Fitelson, and Tentori 2008; Shogenji 2012). that is, because the connection between h and e is plausibly one of
probabilistic dependence but not of logical entailment.)

Remarks such as the foregoing have induced some contemporary Bayesian three distinct, classical families of measures, as follows (Crupi, Chater,
theorists to dismiss the notion of confirmation as firmness altogether, and Tentori 2013; Crupi and Tentori 2016; Heckerman 1988; Sprenger and
concluding with I.J. Good (1968, 134) that “if you had P(h ∣ e ∧ k) close Hartmann 2020, Ch. 1):
to unity, but less than P(h ∣ k) , you ought not to say that h was confirmed
by e ” (also see Salmon 1975, 13). Let us follow this suggestion and Theorem 2
proceed to consider the ordinal (and quantitative) notions of relevance Given (P0) and (P1):
i. (P4) holds if and only if CP (h, e ∣ k) is a probability difference

confirmation.
measure, that is, if there exists a strictly increasing function f
3.4 Differences, ratios, and partial entailment such that, for any h, e, k ∈ L and any P ∈ P,
CP (h, e ∣ k) = f [P(h ∣ e ∧ k) − P(h ∣ k)];
Just as with firmness, the ordinal analysis of relevance confirmation can be
ii. (P5) holds if and only if CP (h, e ∣ k) is a probability ratio
characterized axiomatically. With the relevance notion, however, a larger
measure, that is, if there exists a strictly increasing function f
set of options arises. Consider the following principles.
such that, for any h, e, k ∈ L and any P ∈ P,
CP (h, e ∣ k) = f [ P(h∣k) ];
P(h∣e∧k)
(P4) Disjunction of alternative hypotheses
For any e, h1 , h2 , k ∈ L and any P ∈ P, if k ⊨ ¬(h1 ∧ h2 ), then iii. (P6) holds if and only if CP (h, e ∣ k) is a likelihood ratio measure,
CP (h1 , e ∣ k) ⋛ CP (h1 ∨ h2 , e ∣ k) if and only if that is, if there exists a strictly increasing function f such that, for
P(h2 ∣ e ∧ k) ⋛ P(h2 ∣ k). any h, e, k ∈ L and any P ∈ P, CP (h, e ∣ k) = f [ P(e∣¬h∧k) ].
P(e∣h∧k)
(P5) Law of likelihood If a strictly additive behavior (SA above) is imposed, one functional form
For any e, h1 , h2 , k ∈ L and any P ∈ P, CP (h1 , e ∣ k) ⋛ CP (h2 , e ∣ k) is singled out for the quantitative representation of confirmation
if and only if P(e ∣ h1 ∧ k) ⋛ P(e ∣ h2 ∧ k). corresponding to each of the clauses above:
(P6) Modularity (for conditionally independent data) i. DP (h, e ∣ k) = P(h ∣ e ∧ k) − P(h ∣ k);
e1 , e2 , h, k ∈ L P ∈ P,
ii. RP (h, e ∣ k) = log[ ];
For any and any if P(h∣e∧k)
P(e1 ∣ ±h ∧ e2 ∧ k) = P(e1 ∣ ±h ∧ k), then P(h∣k)
iii. LP (h, e ∣ k) = log[ P(e∣¬h∧k) ].
P(e∣h∧k)
CP (h, e1 ∣ e2 ∧ k) = CP (h, e1 ∣ k).
All the above conditions occur more or less widely in the literature (see (The bases of the logarithms are assumed to be strictly greater than 1.)
Crupi, Chater, and Tentori 2013 and Crupi and Tentori 2016 for references
Before discussing briefly this set of alternative quantitative measures of
and discussion). Interestingly, they’re all pairwise incompatible on the
relevance confirmation, we will address one further related issue. It is a
background of the Formality and the Final Probability principles (P0 and
long-standing idea, going back to Carnap at least, that confirmation theory
P1 above). Indeed, they sort out the relevance notion of confirmation into

⎧ P(h ∣ e ∧ k) − P(h ∣ k)
should yield an inductive logic that is analogous to classical deductive
⎪ if P(h ∣ e ∧ k) ≥ P(h ∣ k)
logic in some suitable sense, thus providing a theory of partial entailment, ⎪ 1 − P(h ∣ k)
Z(h, e ∣ k) = ⎨
and partial refutation. Now, the deductive-logical notions of entailment ⎪ P(h ∣ e ∧ k) − P(h ∣ k)
⎪ if P(h ∣ e ∧ k) < P(h ∣ k)
⎩
and refutation (contradiction) exhibit the following well-known properties:
P(h ∣ k)
Contraposition of entailment
So, despite some pessimistic suggestions (see, e.g., Hawthorne 2018, and
Entailment is contrapositive, but not commutative. That is, it holds
that e entails h (e ⊨ h) if and only if ¬h entails ¬e (¬h ⊨ ¬e), while it
the discussion in Crupi and Tentori 2013), a neat confirmation-theoretic
does not hold that e entails h if and only if h entails e (h ⊨ e).

generalization of logical entailment (and refutation) is possible after all.
Interestingly, relative distance measures can be additive, but only for
Commutativity of refutation uniform pairs of arguments – both confirmatory or both disconfirmatory
Refutation, on the contrary, is commutative, but not contrapositive. (see Milne 2014, p. 259). (Note: Crupi, Tentori, and Gonzalez 2007;
That is, it holds that e refutes h (e ⊨ ¬h) if and only if h refutes e Crupi, Festa, and Buttasi 2010; and Crupi and Tentori 2013, 2014, provide
(h ⊨ ¬e), while it does not hold that e refutes h if and only if ¬h further discussions of the properties of relative distance measures and their
refutes ¬e (¬h ⊨ ¬¬e). intuitive motivations. Also see Mura 2008 for a related analysis.)
The confirmation-theoretic counterparts are fairly straightforward: The plurality of alternative probabilistic measures of relevance
confirmation has prompted some scholars to be skeptical or dismissive of
(P7) Contraposition of confirmation the prospects for a quantitative theory of confirmation (see, e.g., Howson
For any e, h, k ∈ L and any P ∈ P, if e relevance-confirms h relative 2000, 184–185, and Kyburg and Teng 2001, 98 ff.). However, as we will
to k, then CP (h, e ∣ k) = CP (¬e, ¬h ∣ k). see shortly, quantitative analyses of relevance confirmation have proved
important for handling a number of puzzles and issues that plagued
(P8) Commutativity of disconfirmation competing approaches. Moreover, various arguments in the philosophy of
For any e, h, k ∈ L and any P ∈ P, if e relevance-disconfirms h science and beyond have been shown to depend critically (and sometimes
relative to k, then CP (h, e ∣ k) = CP (e, h ∣ k). unwittingly) on the choice of one confirmation measure (or some of them)
rather than others (see Festa and Cevolani 2017, Fitelson 1999, Brössel
The following can then be proven (Crupi and Tentori 2013):
2013, Glass 2013, Roche and Shogenji 2014, Rusconi et al. 2014, and van
Theorem 3 Enk 2014).
Given (P0) and (P1), (P7) and (P8) hold if and only if CP (h, e ∣ k) is a
Recently, arguments have been offered by Huber (2008b) in favor of D , by
relative distance measure, that is, if there exists a strictly increasing
Park (2014), Pruss (2014), and Vassend (2015) in favor of L (also see
function f such that, for any h, e, k ∈ L and any P ∈ P,
CP (h, e ∣ k) = f [Z(h, e ∣ k)], where:
Morey, Romeijn, and Rouder 2016 for an important connection with
statistics), and by Crupi and Tentori (2010) in favor of Z . Hájek and Joyce
⎧ P(h ∣ e ∧ k) − P(h ∣ k)
⎪
⎪
Z(h, e ∣ k) = ⎨
(2008, 123), on the other hand, have seen different measures as possibly Formally, it is fairly simple to show that (SP) characterizes relevance
capturing “distinct, complementary notions of evidential support” (also confirmation (see, e.g., Crupi, Festa, and Buttasi 2010, 80; Hájek and
see Schlosshauer and Wheeler 2011, Sprenger and Hartmann 2020, Ch.1, Joyce 2008, 123), but the philosophical import of this result is nonetheless
and Steel 2007 for tempered forms of pluralism). The case of measure R remarkable. For illustrative purposes, it is useful to assume the
deserves some more specific comments, however. Following Fitelson endorsement of the principle of total evidence (TE) as a default position
(2007), one could see R as conveying key tenets of so-called for the Bayesian. This means that P is assumed to represent actual degrees
“likelihoodist” position about evidential reasoning (see Royall 1997 for a of belief of a rational agent, that is, given all the background information
classical statement, and Chandler 2013 and Sober 1990 for consonant available. Then, by clause (i) of (SP), we have that the occurrence of e , a
arguments and inclinations). There seems to be some consensus, however, consequence of h ∧ k (but not of k alone), confirms h relative to k
that compelling objections can be raised against the adequacy of R as a provided that e was initially uncertain to some degree (even given k ). In
proper measure of relevance confirmation (see, in particular, Crupi, Festa, other words: e must have been predicted on the basis of h ∧ k. Moreover,
and Buttasi 2010, 85–86; Eells and Fitelson 2002; Gillies 1986, 112; and again by (i), the confirmatory impact will be stronger the more surprising
compare Milne 1996 with Milne 2010, Other Internet Resources). In what (unlikely) the evidence was unless h was conjoined to k . So, under TE,
follows, too, it will be convenient to restrict our discussion to D, L and Z relevance confirmation turns out to embed a squarely predictivist version
as candidate measures. All the results to be presented below are invariant of hypothetico-deductivism! As we know, this neutralizes the charge of
for whatever choice among these three options, and across ordinal underdetermination, yet it comes at the usual cost, namely, the old
equivalence with each of them (but those results do not always extend to evidence problem. In fact, if TE is in force, then clause (ii) of (SP) implies
measures ordinally equivalent to R ). that no statement that is known to be true (thus assigned probability 1) can
ever have confirmatory import.
3.5 New evidence, old evidence, and total evidence
Interestingly, the Bayesian predictivist has an escape (neatly anticipated,
Let us go back to a classical HD case, where the (consistent) conjunction and criticized, by Glymour 1980a, 91–92). Consider Einstein and Mercury
h ∧ k (but not k alone) entails e. The following can be proven: once again. As effectively pointed out by Norton (2011a, 7), Einstein was
extremely careful to emphasize that the precession phenomenon had been
Surprising prediction theorem (SP) derived “without having to posit any special [auxiliary] hypotheses at all”.
For any e, h, k ∈ L and any P ∈ P such that h ∧ k ⊨ e and k ⊭ e : Why? Well, presumably because if one had allowed herself to arbitrarily
devise ad hoc auxiliaries (within k , in our notation) then one could have
i. if P(e ∣ k) < 1, then e relevance-confirms h relative to k and
been pretty much certain in advance to find a way to get Mercury’s data
CP (h, e ∣ k) is a decreasing function of P(e ∣ k);
right (remember: that’s the lesson of the underdetermination theorem). But
ii. if P(e ∣ k) = 1, then e is relevance-neutral for h relative to k.
getting those data right with auxiliaries k that were not thus adjusted—that
would have been a natural consequence had the theory of general
relativity been true and it would have been surprising otherwise.

Arguably, this line of argument exploits much of the use-novelty idea handle this problem. First, unlike HD, the Bayesian framework has the
within a predictivist framework. The crucial points are (i) that the formal resources to characterize the auxiliaries themselves as more or less
evidence implied is not a verified empirical statement e but the logical fact likely and thus their adoption as relatively safe or suspicious (the standard
that h ∧ k entails e , and (ii) that the existence of this connection of Bayesian treatment of auxiliary hypotheses is developed along these lines
entailment was not to be obviously anticipated at all, precisely because in Dorling 1979 and Howson and Urbach 2006, 92–102, and it is critically
h ∧ k and e are such that the latter did not serve as a constraint to specify discussed in Rowbottom 2010, Strevens 2001, and Worrall 1993; also see
the former. On these conditions, it seems that h can be confirmed by this Christensen 1997 for an important analysis of related issues). Second, one
kind of “second-order” (logical) evidence in line with (SP) while TE is has to provide indications as to how TE should be relaxed. Non-TE
concurrently preserved. Bayesians of the impermissivist strand often suggest that objective
likelihood values concerning the outcome e —P(e ∣ h ∧ k) —can be
At least two main problems arise, however. The first one is more technical specified for the competing hypotheses at issue quite apart from the fact
in nature. Modelling rational uncertainty concerning logical facts (such as that e may have already occurred. Such values would typically be diverse
h ∧ k ⊨ e) by probabilistic means is no trivial task. Garber (1983) put for different hypotheses (thus mathematically implying P(e ∣ k) < 1) and
forward an influential proposal, but doubts have been raised that it might serve as a basis to capture formally the confirmatory impact of e (see
not be well-behaved (e.g., van Fraassen 1988; a careful survey with further Hawthorne 2005 for an argument along these lines). Permissivists, on the
references can be found in Eva and Hartmann forthcoming). Second, and other hand, can not coherently rely on these considerations to articulate a
more substantially, this solution of the old evidence problem can be non-TE position. They must invoke counterfactual degrees of belief
charged of being an elusive change of the subject: for it was Mercury’s instead, suggesting that P should be reconstructed as representing the
data, not anything else, that had to be recovered as having confirmed (and beliefs that the agent would have, had she not known that e was true (see
still confirming, some would add) Einstein’s theory. That’s the kind of Howson 1991 for a statement and discussion, and Sprenger 2015 for an
judgment that confirmation theory must capture, and which remains original recent variant; also see Jeffrey 1995 and Wagner 2001 for relevant
unattainable for the predictivist Bayesian. (Earman 1992, 131 voiced this technical results, and Steele and Werndl 2013 for an intriguing case-study
complaint forcefully. Hints for a possible rejoinder appear in Eells’s 1990 from climate science).
thorough discussion; see also Skyrms 1983.)
Bayesians that are unconvinced by the predictivist position are naturally

3.6 Paradoxes probabilified and other elucidations
led to dismiss TE and allow for the assignment of initial probabilities
The theory of Bayesian confirmation as relevance indicates when and why
the HD idea works: if h ∧ k (but not k ) entails e , then h is relevance-
lower than 1 even to statements that were known all along. Of course, this
brings the underdetermination problem back, for now k can still be
confirmed by e (relative to k ) because the latter increases the probability
concocted ad hoc to have known evidence e following from h ∧ k and
of the former—provided that P(e ∣ k) < 1. Admittedly, the meaning of the
moreover P(e ∣ k) < 1 is not prevented by TE anymore, thus potentially
latter proviso partly depends on how one handles the problem of old
licencing arbitrary confirmation relations. Two moves can be combined to

evidence. Yet it seems legitimate to say that Bayesian relevance confirmation directly implies that e = black(a) confirms h given
confirmation (unlike the firmness view) retains a key point of ordinary k = raven(a) and e∗ = ¬raven(a) confirms h given k ∗ = ¬black(a)
scientific practice which is embedded in HD and yields further elements of (provided, as we know, that P(e ∣ k) < 1 and P(e∗ ∣ k ∗ ) < 1). That’s
clarification. Consider the following illustration. because h ∧ k ⊨ e and h ∧ k ∗ ⊨ e∗ . But of course, to have h confirmed,
sampling ravens and finding a black one is intuitively more significant
(e1 ) tigers carry the ND1 gene than failing to find a raven while sampling the enormous set of the non-
(e2 ) elephants carry the ND1 gene black objects. That is, it seems, because the latter is very likely to obtain
(e∗2 ) lions carry the ND1 gene anyway, whether or not h is true, so that P(e∗ ∣ k ∗ ) is actually quite close
(h) all mammals carry the ND1 gene to unity. Accordingly, (SP) implies that h is indeed more strongly
confirmed by black(a) given raven(a) than it is by ¬raven(a) given
Qualitative confirmation theories comply with the idea that h is confirmed
¬black(a)—that is, CP (h, e ∣ k) > CP (h, e∗ ∣ k ∗ ) —as long as the
both by e1 ∧ e2 and by e1 ∧ e∗2 . In the HD case, it is clear that h entails
assumption P(e ∣ k) < P(e∗ ∣ k ∗ ) applies.
both conjunctions, given of course k stating that tigers, lions, and
elephants are all mammals (an Hempelian account could also be given What then if the sampling in not constrained (k = ⊤) and the evidence
easily). Bayesian relevance confirmation unequivocally yields the same now amounts to the finding of a black raven, e = raven(a) ∧ black(a),
qualitative verdict. There is more, however. Presumably, one might also versus a non-black non-raven, e∗ = ¬black(a) ∧ ¬raven(a)? We’ve
want to say that h is more strongly confirmed by e1 ∧ e2 than by e1 ∧ e∗2 , already seen that, for either Hempelian or HD-confirmation, e and e∗ are
because the former offers a more varied and diverse body of positive on a par: both Hempel-confirm h, none HD-confirms it. In the former case,
evidence (interestingly, on experimental investigation, this pattern prevails the original Hempelian version of the ravens paradox immediately arises;
in most people’s judgment, including children, see Lo et al. 2002). Indeed, in the latter, it is avoided, but at a cost: e is declared flatly irrelevant for
the variety of evidence is a fairly central issue in the analysis of h—a bit of a radical move. Can the Bayesian do any better? Quite so.
confirmation (see, e.g., Bovens and Hartmann 2002, Schlosshauer and Consider the following conditions:
Wheeler 2011, and Viale and Osherson 2000). In the illustrative case
above, higher variety is readily captured by lower probability: it just seems i. P[raven(a) ∣ h] = P[raven(a)] > 0
a priori less likely that species as diverse as tigers and elephants share ii. P[¬raven(a) ∧ black(a) ∣ h] = P[¬raven(a) ∧ black(a)]
some unspecified genetic trait as compared to tigers and lions, that is,
P(e1 ∧ e2 ∣ k) < P(e1 ∧ e∗2 ∣ k). By (SP) above, then, one immediately Roughly, (i) says that the size of the ravens population does not depend on
gets from the relevance confirmation view the sound implication that their color (in fact, on h), and (ii) that the size of the population of black
CP (h, e1 ∧ e2 ∣ k) > CP (h, e1 ∧ e∗2 ∣ k). non-raven objects also does not depend on the color of ravens. Note that
both (i) and (ii) seem fairly sound as far as our best understanding of our
Principle (SP) is also of much use in the ravens problem. Posit actual world is concerned. It is easy to show that, in relevance-
h = ∀x(raven(x) → black(x)) once again. Just as HD, Bayesian relevance confirmation terms, (i) and (ii) are sufficient to imply that

e = raven(a) ∧ black(a), but not e∗ = ¬raven(a) ∧ ¬black(a), confirms the implication that CP (h, e ∣ k) is positive while CP (h∗ , e ∣ k) is not
h, that is CP (h, e) > CP (h, e∗ ) = 0 (this observation is due to Mat clashes with each of the following, and proves them unduly restrictive: the
Coakley). So the Bayesian relevance approach to confirmation can make a Special Consequence Condition (SCC), the Predictive Inference Condition
principled difference between e and e∗ in both ordinal and qualitative (PIC), and the Consistency Condition (Cons). Note that these principles
terms. (A much broader analysis is provided by Fitelson and Hawthorne were all evaded by HD-confirmation, but all implied by confirmation as
2010, Hawthorne and Fitelson 2010 [Other Internet Resources]. Notably, firmness (see above).
their results include the full specification of the sufficient and necessary
conditions for the main inequality CP (h, e) > CP (h, e∗ ) .) At the same time, the most compelling features of F-confirmation, which
the HD model was unable to capture, are retained by confirmation as
In general, Bayesian (relevance) confirmation theory implies that the relevance. In fact, all our measures of relevance confirmation (D, L , and
evidential import of an instance of some generalization will often depend Z ) entail the ordinal extension of the Entailment Condition (EC) as well as
on the credence structure, and relies on its formal representation, P, as a CP (h, e ∣ k) = −CP (¬h, e ∣ k) and thereby Confirmation Complementarity
tool for more systematic analyses. Consider another instructive example. in all of its forms (qualitative, ordinal, and quantitative). Moreover, the
Assume that a denotes some company from some (otherwise unspecified) Bayesian confirmation theorist of either the firmness or the relevance
sector of the economy, and label the latter predicate S . So, k = Sa. You are strand can avail herself of the same quantitative strategy of “damage
informed that a increased revenues in 2019, represented as e = Ra . Does control” for the main specific paradox of HD-confirmation, i.e., the
this confirm h = ∀x(Sx → Rx) ? It does, at least to some degree, one irrelevant conjunction problem. (See statement (CIC) above, and Crupi
would say. For an expansion of the whole sector (recall that you have no and Tentori 2010, Fitelson 2002. Also see Chandler 2007 for criticism, and
clue what this is) surely would account for the data. That’s a Moretti 2006 for a related debate.)
straightforward HD kind of reasoning (and a suitable Hempelian
counterpart reconstruction would concur). But does e also confirm We’re left with one last issue to conclude our discussion, to wit, the blite
h∗ = Sb → Rb for some further company b ? Well, another obvious paradox. Recall that blite is so defined:
account of the data e would be that company a has gained market shares at
blite(x) ≡ (e xt≤T (x) → black(x)) ∧ (¬ext≤T (x) → white(x)).
the expenses of some competitor, so that e might well seem to support
¬h∗ , if anything (the revenues example is inspired by a remark in Blok, As always heretofore, we posit h = ∀x(raven(x) → black(x)),
Medin, and Osherson 2007, 1362). h∗ = ∀x(raven(x) → blite(x)). We then consider the set up where
k = raven(a) ∧ ext≤T (a), e = black(a), and P(e ∣ k) < 1. Various authors
It can be shown that the Bayesian notion of relevance confirmation allows
have noted that, with Bayesian relevance confirmation, one has that
for this pattern of judgments, because (given k ) evidence e above
P(h ∣ k) > P(h∗ ∣ k) is sufficient to imply that CP (h, e ∣ k) > CP (h∗ , e ∣ k)
increases the probability of h but may well have the opposite effect on h∗
(see Gaifman 1979, 127–128; Sober 1994, 229–230; and Fitelson 2008,
(see Sober 1994 for important remarks along similar lines). Notably, h
131). So, as long as the black hypothesis is perceived as initially more
entails h∗ by plain instantiation, and so contradicts ¬h∗ . As a consequence,

credible than its blite counterpart, the former will be more strongly accommodation/>.
confirmed than the latter. Of course, P(h ∣ k) > P(h∗ ∣ k) is an entirely Betz, G., 2013, “Revamping Hypothetico-Deductivism: A Dialectic
commonsensical assumption, yet these same authors have generally, and Account of Confirmation”, Erkenntnis, 78: 991–1009.
quite understandably, failed to see this result as philosophically Bird, A., 2010, “Eliminative Abduction—Examples from Medicine”,
illuminating. Lacking some interesting, non-question-begging story as to Studies in History and Philosophy of Science, 41: 345–352.
why that inequality should obtain, no solution of the paradox seems to Blok, S.V., D.L. Medin, and D. Osherson, 2007, “Induction as Conditional
emerge. More modestly, one could point out that a measure of relevance Probability Judgment”, Memory & Cognition, 35: 1353–1364.
confirmation CP (h, e ∣ k) implies (i) and (ii) below. Bovens, L. and S. Hartmann, 2002, “Bayesian Networks and the Problem
of Unreliable Instruments”, Philosophy of Science, 69: 29–72.
i. Necessarily (that is, for any P ∈ P), e confirms h relative to k. Brössel, P., 2013, “The Problem of Measure Sensitivity Redux”,
ii. Possibly (that is, for some P ∈ P), each one of the following obtains: Philosophy of Science, 80: 378–397.
e confirms that a raven will be black if examined after T , that is, Carnap, R., 1950/1962, Logical Foundations of Probability, Chicago:
(raven(b) ∧ ¬ext≤T (b)) → black(b), relative to k; and University of Chicago Press.
e does not confirm that a raven will be white if examined after Chandler, J., 2007, “Solving the Tacking Problem with Contrast Classes”,
T , that is, (raven(b) ∧ ¬ext≤T (b)) → white(b), relative to k. British Journal for the Philosophy of Science, 58: 489–502.
–––, 2013, “Contrastive Confirmation: Some Competing Accounts”,
Without a doubt, (i) and (ii) fall far short of a satisfactory solution of the
Synthese, 190: 129–138.
blite paradox. Yet it seems at least a legitimate minimal requirement for a
Christensen, D., 1997, “What Is Relative Confirmation?”, Noûs, 3: 370–
compelling solution (if any exists) that it implies both. It is then of interest
384.
to note that confirmation as firmness is inconsistent with (i), while
Cleland, C.E., 2011, “Prediction and Explanation in Historical Natural
Hempelian and HD-confirmation are inconsistent with (ii).
Science”, British Journal for the Philosophy of Science, 62: 551–582.
Craig, W., 1957, “The Uses of the Herbrand-Gentzen Theorem in Relating
Bibliography
Model Theory and Proof Theory”, Journal of Symbolic Logic, 3:
Achinstein, P. (ed.), 2005, Scientific Evidence: Philosophical Theories and 269–285.
Applications, Baltimore: John Hopkins University Press. Crupi, V., 2015, “Inductive Logic”, Journal of Philosophical Logic, 44
Baker, A., 2003, “Quantitative Parsimony and Explanatory Power”, (40th anniversary issue): 641–650.
British Journal for the Philosophy of Science, 54: 245–259. Crupi, V. and K. Tentori, 2010, “Irrelevant Conjunction: Statement and
Barnes, E.C., 2018, “Prediction versus Accommodation”, The Stanford Solution of a New Paradox”, Philosophy of Science, 77: 1–13.
Encyclopedia of Philosophy (Fall 2018 Edition), E.N. Zalta (ed.), –––, 2013, “Confirmation as Partial Entailment: A Representation
URL = Theorem in Inductive Logic”, Journal of Applied Logic, 11: 364–372
<https://plato.stanford.edu/archives/fall2018/entries/prediction- [Erratum in Journal of Applied Logic, 12: 230–231].

–––, 2014, “Measuring Information and Confirmation”, Studies in the (NJ): Princeton University Press, 1991.
History and Philosophy of Science, 47: 81–90. Earman, J., 1992, Bayes or Bust?, Cambridge (MA): MIT Press.
–––, 2016, “Confirmation Theory”, in A. Hájek and C. Hitchcock (eds.), ––– (ed.), 1983, Minnesota Studies in Philosophy of Science, Vol. 10:
Oxford Handbook of Philosophy and Probability, Oxford: Oxford Testing Scientific Theories, Minneapolis: University of Minnesota
University Press, pp. 650–665. Press.
Crupi, V., N. Chater, and K. Tentori, 2013, “New Axioms for Probability Earman, J. and C. Glymour, 1988, “What Revisions Does Bootstrap
and Likelihood Ratio Measures”, British Journal for the Philosophy Testing Need?”, Philosophy of Science, 55: 260–264.
of Science, 64: 189–204. Earman, J. and M. Janssen, 1993, “Einstein’s Explanation of the Motion of
Crupi, V., R. Festa, and C. Buttasi, 2010, “Towards a Grammar of Mercury’s Perihelion”, in J. Earman, M. Janssen, and J. Norton
Bayesian Confirmation”, in M. Suárez, M. Dorato, and M. Rédei (eds.), The Attraction of Gravity: New Studies in the History of
(eds.), Epistemology and Methodology of Science, Dordrecht: General Relativity, Boston: Birkhäuser, pp. 129–172.
Springer, pp. 73–93. Earman, J. and W.C. Salmon, 1992, “The Confirmation of Scientific
Crupi, V., B. Fitelson, and K. Tentori, 2008, “Probability, Confirmation, Hypotheses”, in M.H. Salmon et al., Introduction to the Philosophy of
and the Conjunction Fallacy”, Thinking & Reasoning, 14: 182–199. Science, Englewood Cliff: Prentice Hall, pp. 42–103.
Crupi, V., K. Tentori, and M. Gonzalez, 2007, “On Bayesian Measures of Easwaran, K., 2011a, “Bayesianism I: Introduction and Arguments in
Evidential Support: Theoretical and Empirical Issues”, Philosophy of Favor”, Philosophy Compass, 6: 312–320.
Science 74: 229–52. –––, 2011b, “Bayesianism II: Criticisms and Applications”, Philosophy
Dawes, R.M., 2001, Everyday Irrationality, Boulder (CO): Westview. Compass, 6: 321–332.
De Finetti, B., 2008, Philosophical Lectures on Probability (edited by A. Eells, E., 1990, “Bayesian Problems of Old Evidence”, in Savage, 1990,
Mura), Dordrecht: Springer. pp. 205–223.
Dorling, J., 1979, “Bayesian Personalism, the Methodology of Scientific Eells, E. and B. Fitelson, 2002, “Symmetries and Asymmetries in
Research Programmes, and Duhem’s Problem”, Studies in History Evidential Support”, Philosophical Studies, 107: 129–42.
and Philosophy of Science, 10: 177–187. Eva, B. and S. Hartmann, forthcoming, “On the Origins of Old Evidence”,
Douglas, H. and P.D. Magnus, 2013, “Why Novel Prediction Matters”, Australasian Journal of Philosophy, first online 14 October 2019.
Studies in History and Philosophy of Science, 44: 580–589. doi:10.1080/00048402.2019.1658210
Douven, I., 2017, “Abduction”, in E.N. Zalta (ed.), The Stanford Festa, R., 1999, “Bayesian Confirmation”, in M. Galavotti and A. Pagnini
Encyclopedia of Philosophy (Summer 2017 Edition), URL = (eds.), Experience, Reality, and Scientific Explanation, Dordrecht:
<https://plato.stanford.edu/archives/sum2017/entries/abduction/>. Kluwer, pp. 55–87.
Douven, I. and W. Meijs, 2006, “Bootstrap Confirmation Made Festa, R. and G. Cevolani, 2017, “Unfolding the Grammar of Bayesian
Quantitative”, Synthese, 149: 97–132. Confirmation: Likelihood and Anti-Likelihood Principles”,
Duhem, P., 1906, The Aim and Structure of Physical Theory, Princeton Philosophy of Science, 84: 56–81.

–––, 2003, “Induction, Probability, and Bayesian Epistemology”, Poznan –––, 1998, “Hypothetico-Deductivism: The Current State of Play; The
Studies in the Philosophy of Science and the Humanities, 80: 251– Criterion of Empirical Significance: Endgame”, Erkenntnis, 49: 1–20.
284. –––, 2005, “Hypothetico-Deductivism: Incomplete But Not Hopeless”,
Fitelson, B., 1999, “The Plurality of Bayesian Measures of Confirmation Erkenntnis, 63: 139–147.
and the Problem of Measure Sensitivity”, Philosophy of Science, 66: Giere, R.N., 1983, “Testing Theoretical Hypotheses”, in Earman, 1983,
S362–78. pp. 269–298.
–––, 2002, “Putting the Irrelevance Back into the Problem of Irrelevant –––, 1979, Understanding Scientific Reasoning, Belmont (CA): Thomson /
Conjunction”, Philosophy of Science, 69: 611–622. Wadsworth, 2006.
–––, 2007, “Likelihoodism, Bayesianism, and Relational Confirmation”, Gillies, D., 1986, “In Defense of the Popper-Miller Argument”,
Synthese 156: 473–89. Philosophy of Science, 53: 110–113.
–––, 2008, “Goodman’s New Riddle”, Journal of Philosophical Logic, 37: –––, 1989, “Non-Bayesian Confirmation Theory and the Principle of
613–643. Explanatory Surplus”, in A. Fine and J. Leplin (eds.), Proceedings of
Fitelson, B. and J. Hawthorne, 2010, “How Bayesian Confirmation Theory the 1988 Biennial Meeting of Philosophy of Science Association, Vol.
Handles the Paradox of the Ravens”, in E. Eells and J. Fetzer (eds.), 2, East Lansing (MI): Philosophy of Science Association, pp. 381–
The Place of Probability in Science, Dordrecht: Springer, pp. 247– 392.
276. –––, 2000, Philosophical Theories of Probability, London: Routledge.
Forster, M., 2007, “A Philosopher’s Guide to Empirical Success”, Glass, D.H., 2013, “Confirmation Measures of Association Rule
Philosophy of Science, 74: 588–600. Interestingness”, Knowledge-Based Systems, 44: 65–77.
Forster, M. and E. Sober, 1994, “How to Tell When Simpler, More Glymour, C., 1980a, Theory and Evidence, Princeton (NJ): Princeton
Unified, or Less Ad Hoc Theories Will Provide More Accurate University Press.
Predictions”, British Journal for the Philosophy of Science, 45: 1–35. –––, 1980b, “Hypothetico-Deductivism Is Hopeless”, Philosophy of
Franklin, J., 2001, “Resurrecting Logical Probability”, Erkenntnis, 55: Science, 47: 322–325.
277–305. Good, I.J., 1967, “The White Shoe Is a Red Herring”, British Journal for
Gabbay, D., S. Hartmann, and J. Woods (eds.), 2011, Handbook of the the Philosophy of Science, 17: 322.
History of Logic, Vol. 10: Inductive Logic, Amsterdam: Elsevier. –––, 1968, “Corroboration, Explanation, Evolving Probabilities,
Gaifman, H., 1979, “Subjective Probability, Natural Predicates, and Simplicity, and a Sharpened Razor”, British Journal for the
Hempel’s Ravens”, Erkenntnis, 14: 105–147. Philosophy of Science, 19: 123–43.
Garber, D., 1983, “Old Evidence and Logical Omniscience”, in Earman, –––, 1971, “46656 Varieties of Bayesians”, American Statistician, 25 (5):
1983, pp. 99–131. 62–63.
Gemes, K. 1993, “Hypothetico-Deductivism, Content and the Natural Goodman, N., 1955, Fact, Fiction, and Forecast, Cambridge (MA):
Axiomatization of Theories”, Philosophy of Science, 60: 477–487. Harvard University Press, 1983.

Guala, F., 2012, “Experimentation in Economics”, in D. Gabbay, P. Hesse, M. 1975, “Bayesian Methods and the Initial Probabilities of
Thagard, J. Woods, and U. Mäki (eds.), Handbook of the Philosophy Theories”, in Maxwell and Anderson, 1975, pp. 50–105.
of Science: Philosophy of Economics, Amsterdam: Elsevier, pp. 597– Hitchcock, C. and E. Sober, 2004, “Prediction versus Accommodation and
640. the Risk of Overfitting”, British Journal for the Philosophy of
Hájek, A., 2008. “Arguments For—Or Against—Probabilism?”, British Science, 55: 1–34.
Journal for the Philosophy of Science, 59: 793–819. Horvitz, E. and D. Heckerman, 1986, “The Inconsistent Use of Measures
–––, 2019, “Interpretations of Probability”, E.N. Zalta (ed.), The Stanford of Certainty in Artificial Intelligence Research”, in L.N. Kanal and
Encyclopedia of Philosophy (Fall 2019 Edition), URL = J.F. Lemmer (eds.), Uncertainty in Artificial Intelligence,
<https://plato.stanford.edu/archives/fall2019/entries/probability- Amsterdam: North-Holland, pp. 137–151.
interpret/>. Horwich, P., 1983, “Explanations of Irrelevance”, in Earman, 1983, pp.
Hájek, A. and J. Joyce, 2008, “Confirmation”, in S. Psillos and M. Curd 55–65.
(eds.), Routledge Companion to the Philosophy of Science, New Hosiasson-Lindenbaum, J., 1940, “On Confirmation”, Journal of Symbolic
York: Routledge, pp. 115–29. Logic, 5: 133–148.
Hawthorne, J., 2005, “Degrees-of-Belief and Degrees-of-Support: Why Howson, C., 1990, “Fitting Theory to the Facts: Probably Not Such a Bad
Bayesians Need Both Notions”, Mind, 114: 277–320. Idea After All”, in Savage, 1990, pp. 224–244.
–––, 2011, “Confirmation Theory”, in D. Gabbay, P. Thagard, J. Woods, –––, 1991, “The ‘Old Evidence’ Problem”, British Journal for the
P.S. Bandyopadhyay, and M. Forster (eds.), Handbook of the Philosophy of Science, 42: 547–555.
Philosophy of Science: Philosophy of Statistics, Dordrecht: Elsevier, –––, 2000, Hume’s Problem: Induction and the Justification of Belief,
pp. 333–389. Oxford: Oxford University Press.
–––, 2018, “Inductive Logic”, in E.N. Zalta (ed.), The Stanford Howson, C. and P. Urbach, 2006, Scientific Reasoning. The Bayesian
Encyclopedia of Philosophy (Spring 2018 Edition), URL = Approach, La Salle (IL): Open Court.
<https://plato.stanford.edu/archives/spr2018/entries/logic- Huber, F., 2008a, “Hempel’s Logic of Confirmation”, Philosophical
inductive/>. Studies, 139: 181–189.
Heckerman, D., 1988, “An Axiomatic Framework for Belief Updates”, in –––, 2008b, “Assessing Theories, Bayes Style”, Synthese, 161: 89–118.
J.F. Lemmer and L.N. Kanal (eds.), Uncertainty in Artificial Huber, F. and C. Schmidt-Petri, 2009, Degrees of Belief, Dordrecht:
Intelligence 2, Amsterdam: North-Holland, pp. 11–22. Springer.
Hempel, C.G., 1937, “Le problème de la vérité”, Theoria, 3: 206–246. Janssen, M., 2003, “COI Stories: Explanation and Evidence in the History
–––, 1943, “A Purely Syntactical Definition of Confirmation”, Journal of of Science”, Perspectives on Science, 10: 457–522.
Symbolic Logic, 8: 122–143. Jeffrey, R., 1995, “Probability Reparation: The Problem of New
–––, 1945, “Studies in the Logic of Confirmation”, Mind, 54: 1–26, 97– Explanation”, Philosophical Studies, 77: 97–101.
121. Joyce, J., 2019, “Bayes’ Theorem”, in E.N. Zalta (ed.), The Stanford

Encyclopedia of Philosophy (Spring 2019 Edition), URL = The Methodology of Scientific Research Programmes, Cambridge:
<https://plato.stanford.edu/archives/spr2019/entries/bayes-theorem/>. Cambridge University Press, 1978, pp. 168–192).
–––, 2011, “The Development of Subjective Bayesianism”, in Gabbay, Lange, M., 2011, “Hume and the Problem of Induction”, in Gabbay,
Hartmann, and Woods, 2011, pp. 415–476. Hartmann, and Woods, 2011, pp. 43–91.
Kelly, K., 2007, “A New Solution to the Puzzle of Simplicity”, Philosophy Laudan, L., 1990, “Demystifying Underdetermination”, in Savage, 1990,
of Science, 74: 561–573. pp. 267–297.
–––, 2008, “Ockham’s Razor, Truth, and Information”, in D. Gabbay, P. –––, 1997, “How About Bust? Factoring Explanatory Power Back into
Thagard, J. Woods, P. Adriaans, and J. van Benthem (eds.), Theory Evaluation”, Philosophy of Science, 64: 206–216.
Handbook of the Philosophy of Science: Philosophy of Information, Laudan, L. and J. Leplin, 1991, “Empirical Equivalence and
Dordrecht: Elsevier, pp. 321–360. Underdetermination”, Journal of Philosophy, 88: 449–472.
Kelly, K. and C. Glymour, 2004, “Why Probability Does Not Capture the Levi, I., 2008, “Degrees of Belief”, Journal of Logic and Computation,
Logic of Scientific Justification”, in C. Hitchcock (ed.), 18: 699–719.
Contemporary Debates in the Philosophy of Science, London: Liebman, J., J. Fagan, V. West, and J. Lloyd, 2000, “Capital Attrition:
Blackwell, pp. 94–114. Error Rates in Capital Cases, 1973–1995”, Texas Law Review, 78:
Keynes, J., 1921, A Treatise on Probability, London: Macmillan. 1839–1865.
Kitcher, P., 1993. The Advancement of Science, Oxford: Oxford University Lipton, P., 2000, “Inference to the Best Explanation”, in W.H. Newton-
Press. Smith (ed.), A Companion to the Philosophy of Science, Oxford:
Kopec, M. and M. Titelbaum, 2016, “The Uniqueness Thesis”, The Blackwell, pp. 184–193.
Philosophy Compass, 11: 189–200. –––, 2004, Inference to the Best Explanation, London: Routledge.
Kuhn, T., 1957, The Copernican Revolution, Cambridge (MA): Harvard –––, 2005, “Testing Hypotheses: Prediction and Prejudice”, Science, 307:
University Press. 219–221.
–––, 1962/1970, The Structure of Scientific Revolutions, Chicago: Lo, Y., A. Sides, J. Rozelle, and D.N. Osherson, 2002, “Evidential
University of Chicago Press. Diversity and Premise Probability in Young Children’s Inductive
Kuipers, T., 2000, From Instrumentalism to Constructive Realism, Judgment”, Cognitive Science, 26: 181–206.
Dordrecht: Reidel. Maher, P., 1996, “Subjective and Objective Confirmation”, Philosophy of
Kyburg, H.E. and C.M. Teng, 2001, Uncertain Inference, New York: Science, 63: 149–174.
Cambridge University Press. –––, 2006, “Confirmation Theory”, in D.M. Borchert (ed.), Encyclopedia
Lakatos, I. and E. Zahar, 1975, “Why did Copernicus’ Research of Philosophy (2nd edition), Detroit (MI): Macmillan Reference.
Programme Supersede Ptolemy’s?”, in R.S. Westman (ed.), The –––, 2010, “Explication of Inductive Probability”, Journal of
Copernican Achievement, Berkeley (CA): University of California Philosophical Logic, 39: 593–616.
Press, pp. 354–383 (reprinted in Lakatos, I., Philosophical Papers I: Martens, R., 2009, “Harmony and Simplicity: Aesthetic Virtues and the

Rise of Testability”, Studies in History and Philosophy of Science, Bayes Factors and the Quantification of Statistical Evidence”,
40: 258–266. Journal of Mathematical Psychology, 72: 6–18.
Maxwell, G. and R.M. Anderson, Jr. (eds.), 1975, Minnesota Studies in Mura, A., 2008, “Can Logical Probability Be Viewed as a Measure of
Philosophy of Science, Vol. 6: Induction, Probability, and Degrees of Partial Entailment?”, Logic & Philosophy of Science 6:
Confirmarion, Minneapolis: University of Minnesota Press. 25–33.
Mayo, D., 1991, “Novel Evidence and Severe Tests”, Philosophy of Musgrave, A., 1974, “Logical versus Historical Theories of
Science, 58: 523–553. Confirmation”, British Journal for the Philosophy of Science, 25: 1–
–––, 1996, Error and the Growth of Experimental Knowledge, Chicago: 23.
University of Chicago Press. –––, 2010, “Critical Rationalism, Explanation, and Severe Tests”, in D.
–––, 2010,“Learning from Error: The Theoretical Significance of Mayo and A. Spanos, 2010, pp. 88–112.
Experimental Knowledge”, The Modern Schoolman, 87: 191–217. Nicod, J., 1924, Le problème logique de l’induction, Paris: Alcan. (Engl.
–––, 2014,“Some Surprising Facts about (the Problem of) Surprising transl. “The Logical Problem of Induction”, in Foundations of
Facts”, Studies in the History and Philosophy of Science, 45: 79–86. Geometry and Induction, London: Routledge, 2000.)
Mayo, D. and A. Spanos (eds.), 2010, Error and Inference: Recent Norton, J., 2005, “A Little Survey on Induction”, in P. Achinstein (ed.),
Exchanges on Experimental Reasoning, Reliability, and the Scientific Evidence: Philosophical Theories and Applications,
Objectivity and Rationality of Science, Cambridge and London: Baltimore: John Hopkins University Press, pp. 9–34.
Cambridge University Press. –––, 2011a, “History of Science and the Material Theory of Induction:
Meacham, C.J.G., 2014, “Impermissive Bayesianism”, Erkenntnis, 79: Einstein’s Quanta, Mercury’s Perihelion”, European Journal for
1185–1217. Philosophy of Science, 1: 3–27.
Meehl, P.E., 1990, “Appraising and Amending Theories: The Strategy of –––, 2011b, “Challenges to Bayesian Confirmation Theory”, in D.
Lakatosian Defense and Two Principles That Warrant Using It”, Gabbay, P. Thagard, J. Woods, S. Bandyopadhyay, and M. Forster
Psychological Inquiry, 1: 108–141. (eds.), Handbook of the Philosophy of Science: Philosophy of
Mill, J.S., 1843/1872, A System of Logic, Honolulu: University Press of Statistics, Amsterdam: Elsevier, pp. 391–440.
the Pacific, 2002. Oaksford, M. and N. Chater, 2007, Bayesian Rationality, Oxford:
Milne, P., 1996, “Log[P(h ∣ eb)/P(h ∣ b)] Is the One True Measure of Clarendon Press.
Confirmation”, Philosophy of Science, 63: 21–6. Okasha, S., 2011, “Experiment, Observation, and the Confirmation of
–––, 2014, “Information, Confirmation, and Conditionals”, Journal of Laws”, Analysis, 71: 222–232.
Applied Logic, 12: 252–262. Osherson, D., E.E. Smith, and E. Shafir, 1986, “Some Origins of Belief”,
Moretti, L., 2006, “The Tacking by Disjunction Paradox: Bayesianism Cognition, 24: 197–224.
versus Hypothetico-Deductivism”, Erkenntnis, 64: 115–138. Paris, J., 2011, “Pure Inductive Logic”, in L. Horsten and R. Pettigrew
Morey, R.D., J.-W. Romeijn, and J.N. Rouder, 2016, “The Philosophy of (eds.), The Continuum Companion to Philosophical Logic, London:

Continuum, pp. 428–449. Duhem’s Thesis Revisited”, Synthese, 177: 139–149.

Park, I., 2014, “Confirmation Measures and Collaborative Belief Royall, R., 1997, Statistical Evidence: A Likelihood Paradigm, London:
Updating”, Synthese, 191: 3955–3975. Chapman & Hall.
Peijnenburg, J., 2012, “A Case of Confusing Probability and Rusconi, P., M. Marelli, M. D’Addario, S. Russo, and P. Cherubini, 2014,
Confirmation”, Synthese, 184: 101–107. “Evidence Evaluation: Measure Z Corresponds to Human Utility
Peirce, C.S., 1878, “The Probability of Induction”, in Philosophical Judgments Better than Measure L and Optimal-Experimental-Design
Writings of Peirce (edited by J. Buchler), New York: Dover, pp. 174– Models”, Journal of Experimental Psychology: Learning, Memory,
189. and Cognition, 40: 703–723.
Pettigrew, R., 2016, Accuracy and the Laws of Credence, Oxford: Oxford Russell, B., 1912, The Problems of Philosophy, Oxford: Oxford University
University Press. Press, 1997.
Popper, K.R., 1934/1959, The Logic of Scientific Discovery, London: Salmon, W.C., 1969, “Partial Entailment as a Basis for Inductive Logic”,
Routledge, 2002. in N. Rescher (ed.), Essays in Honor of Carl G. Hempel, Dordrecht:
–––, 1954, “Degree of Corroboration”, British Journal for the Philosophy Reidel, pp. 47–81.
of Science, 5: 143–149. –––, 1975, “Confirmation and Relevance”, in Maxwell and Anderson,
–––, 1963, Conjectures and Refutations, London: Routledge, 2002. 1975, pp. 3–36.
Pruss, A.R., 2014, “Independent Tests and the Log-Likelihood Ratio –––, 2001, “Explanation and Confirmation: A Bayesian Critique of
Measure of Confirmation”, Thought, 3: 124–135. Inference to the Best Explanation”, in G. Hon and S.S. Rakover
Quine, W.v.O., 1951, “Two Dogmas of Empiricism”, in From a Logical (eds.), Explanation: Theoretical Approaches and Applications,
Point of View, Cambridge (MA): Harvard University Press, pp. 20– Dordrecht: Kluwer, pp. 61–91.
46. Savage, C.W. (ed.), 1990, Minnesota Studies in the Philosophy of Science,
–––, 1970, “Natural Kinds”, in N. Rescher (ed.), Essays in Honor of Carl Vol. 14: Scientific Theories, Minneapolis: University of Minnesota
G. Hempel, Dordrecht: Reidel, pp. 41–56. Press.
Quine, W.v.O. and J. Ullian, 1970, The web of Belief, New York: Random Schippers, M., 2017, “A Representation Theorem for Absolute
House. Confirmation”, Philosophy of Science, 84: 82–91.
Riesch, H., 2010, “Simple or Simplistic? Scientists’ Views on Occam’s Schlosshauer, M. and G. Wheeler, 2011, “Focussed Correlation,
Razor”, Theoria, 67: 75–90. Confirmation, and the Jigsaw Puzzle of Variable Evidence”,
Rinard, S., 2014, “A New Bayesian Solution to the Problem of the Philosophy of Science, 78: 376–392.
Ravens”, Philosophy of Science, 81: 81–100. Schurz, G., 1991, “Relevant Deduction”, Erkenntnis, 35: 391–437.
Roche, W. and T. Shogenji, 2014, “Dwindling Confirmation”, Philosophy –––, 1994, “Relevant Deduction and Hypothetico-Deductivism: A Reply
of Science, 81: 114–137. to Gemes”, Erkenntnis, 41: 183–188.
Rowbottom, D.P., 2010, “Corroboration and Auxiliary Hypotheses: –––, 2005, “Bayesian H-D Confirmation and Structuralistic Truthlikeness:

Discussion and Comparison with the Relevant-Element and the Sprenger, J. and S. Hartmann, 2019, Bayesian Philosophy of Science,
Content-Part Approach”, in R. Festa, A. Aliseda, and J. Peijnenburg Oxford: Oxford University Press.
(eds.), Confirmation, Empirical Progress, and Truth Approximation. Steel, D., 2007, “Bayesian Confirmation Theory and the Likelihood
Essays in Debate with Theo Kuipers, Vol. I, Amsterdam: Rodopi, pp. Principle”, Synthese, 156: 55–77.
141–159. Steele, K. and C. Werndl, 2013, “Climate Models, Calibration, and
–––, 2014, “Bayesian Pseudo-Confirmation, Use-Novelty, and Genuine Confirmation”, British Journal for the Philosophy of Science, 64:
Confirmation”, Studies in the History and Philosophy of Science, 45: 609–635.
87–96. Strasser, C. and G.A. Antonelli, 2019, “Non-monotonic Logic”, The
Schwartz, R., 2011, “Goodman and the Demise of Syntactic and Semantic Stanford Encyclopedia of Philosophy (Summer 2019 Edition), E.N.
Models”, in Gabbay, Hartmann, and Woods, 2011, pp. 391–413. Zalta (ed.), URL =
Scorzato, L., 2013, “On the Role of Simplicity in Science”, Synthese, 190: <https://plato.stanford.edu/archives/sum2019/entries/logic-
2867–2895. nonmonotonic/>.
Shogenji, T., 2012, “The Degree of Epistemic Justification and the Strevens, M., 2001, “The Bayesian Treatment of Auxiliary Hypotheses”,
Conjunction Fallacy”, Synthese, 184: 29–48. British Journal for the Philosophy of Science, 52: 515–537.
Skyrms, B., 1983, “Three Ways to Give a Probability Assignment a Tentori, K., V. Crupi, and D. Osherson, 2007, “Determinants of
Memory”, in Earman, 1983, pp. 157–162. Confirmation”, Psychonomic Bulletin & Review, 14: 877–83.
–––, 1987, “Coherence”, in N. Rescher (ed.) Scientific Inquiry in –––, 2010, “Second-order Probability Affects Hypothesis Confirmation”,
Philosophical Perspective, Pittsburgh: University of Pittsburgh Press, Psychonomic Bulletin & Review, 17: 129–34.
pp. 225–242. Thagard, P.R., 1978, “The Best Explanation: Criteria for Theory Choice”,
Sober, E., 1975, Simplicity, Oxford: Clarendon Press. Journal of Philosophy, 75: 76–92.
–––, 1990, “Contrastive Empiricism”, in Savage, 1990, pp. 392–412. Törnebohm, H., 1966, “Two Measures of Evidential Strength”, in J.
–––, 1994, “No Model, No Inference: A Bayesian Primer on the Grue Hintikka and P. Suppes (eds.), Aspects of Inductive Logic,
Problem”, in D. Stalker (ed.), Grue! The New Riddle of Induction, Amsterdam: North-Holland, pp. 81–95.
Chicago (IL): Open Court, pp. 225–240. van Enk, S.J., 2014, “Bayesian Measures of Confirmation from Scoring
Spohn, W., 2012, The Laws of Belief, Oxford: Oxford University Press. Rules”, Philosophy of Science, 81: 101–113.
Sprenger, J., 2011a, “Hempel and the Paradoxes of Confirmation”, in van Fraassen, B.C., 1988, “The Problem of Old Evidence”, in D.F. Austin
Gabbay, Hartmann, and Woods, 2011, pp. 231–260. (ed.), Philosophical Analysis, Dordrecht: Reidel, pp. 153–165.
–––, 2011b, “Hypothetico-Deductive Confirmation”, Philosophy Compass, –––, B.C., 1989, Laws and Symmetry, Oxford: Oxford University Press.
6: 497–508. Varzi, A.C., 2008, “Patterns, Rules, and Inferences”, in J.A. Adler and L.J.
–––, 2015, “A Novel Solution to the Problem of Old Evidence”, Rips (eds.), Reasoning: Studies in Human Inference and Its
Philosophy of Science, 82: 383–401. Foundations, Cambridge: Cambridge University Press, pp. 282–290.

Vassend, O.B., 2015, “Confirmation Measures and Sensitivity”, <https://plato.stanford.edu/archives/win2019/entries/scientific-

Philosophy of Science, 82: 892–904. explanation/>.
Viale, R. and D. Osherson, 2000, “The Diversity Principle and the Little Worrall, J., 1978, “The Ways in Which the Methodology of Scientific
Scientist Hypothesis”, Foundations of Science, 5: 239–253. Research Programmes Improves on Popper’s Methodology”, in G.
Vineberg, S., 2016, “Dutch Book Arguments”, in E.N. Zalta (ed.), The Radnitzky and G. Andersson (eds.), Progress and Rationality in
Stanford Encyclopedia of Philosophy (Spring 2016 Edition), URL = Science, Dordrecht: Reidel, 1978, pp. 45–70.
<https://plato.stanford.edu/archives/spr2016/entries/dutch-book/>. –––, 1982, “Broken Bootstraps”, Erkenntnis, 18: 105–130.
Votsis, I., 2014, “Objectivity in Confirmation: Post Hoc Monsters and –––, 1990, “Scientific Revolutions and Scientific Rationality: The Case of
Novel Predictions”, Studies in the History and Philosophy of Science, the ‘Elderly Holdout’”, in Savage, 1990, pp. 319–354.
45: 70–78. –––, 1993, “Falsification, Rationality, and the Duhem Problem”, in J.
Wagner, C.G., 2001, “Old Evidence and New Explanation III”, Philosophy Earman, G.J. Massey, and N. Rescher (eds.), Philosophical Problems
of Science, 68: S165–S175. of the Internal and External World: Essays on the Philosophy of A.
Weisberg, J., 2015, “You’ve Come a Long Way, Bayesians”, Journal of Grünbaum, Pittsburgh: University of Pittsburgh Press, pp. 329–370.
Philosophical Logic, 44 (The Fortieth Anniversary Issue): 817–834. –––, 1996, “‘Revolution in Permanence’: Popper on Theory-Change in
Whewell, W., 1840/1847, The Philosophy of the Inductive Sciences, Science”, in A. O’Hear (ed.), Karl Popper: Philosophy and
Founded Upon Their History, Charleston (SC): BiblioBazaar, 2011. Problems, Cambridge: Cambridge University Press, pp. 75–102.
Williamson, J., 2011, “An Objective Bayesian Account of Confirmation”, –––, 2010, “Evidence: Philosophy of Science Meets Medicine”, Journal of
in D. Dieks, W.J. Gonzalez, S. Hartmann, T. Uebel, and M. Weber Evaluation in Clinical Practice, 16: 356–362.
(eds.), Explanation, Prediction, and Confirmation, Berlin: Springer, Zabell, S., 2011, “Carnap and the Logic of Induction”, in Gabbay,
2011, pp. 53–81. Hartmann, and Woods, 2011, pp. 265–310.
Williamson, T., 2002, Knowledge and Its Limits, Oxford: Oxford Zahar, E., 1973, “Why Did Einstein’s Programme Supersede Lorentz’s?”,
University Press. British Journal for the Philosophy of Science, 24: 95–123, 223–262.
Winters, B., J. Custer, S.M. Galvagno Jr, E. Colantuoni, S.G. Kapoor, H. Zellner, A., H. Keuzenkamp, and M. McAleer (eds.), 2002, Simplicity,
Lee, V. Goode, K. Robinson, A. Nakhasi, P. Pronovost, and D. Inference, and Modelling: Keeping It Sophisticatedly Simple,
Newman-Toker, 2012, “Diagnostic Errors in the Intensive Care Unit: Cambridge: Cambridge University Press.
A Systematic Review of Autopsy Studies”, BMJ Quality and Safety,
21: 894–902. Academic Tools
Woodward, J., 1983, “Glymour on Theory Confirmation”, Philosophical
Studies, 43: 147–152. How to cite this entry.
–––, 2019, “Scientific Explanation”, The Stanford Encyclopedia of Preview the PDF version of this entry at the Friends of the SEP
Society.
Philosophy (Winter 2019 Edition), E.N. Zalta (ed.), URL=
Look up this entry topic at the Internet Philosophy Ontology

Confirmation
Project (InPhO).
Enhanced bibliography for this entry at PhilPapers, with links
to its database.
Other Internet Resources

Hawthorne, J. and B. Fitelson, 2010, An Even Better Solution to the
Paradox of the Ravens, manuscript available online (PDF).
Milne, P., 2010, Measuring Confirmation (PDF), slides for a talk at
the 7th Annual Formal Epistemology Workshop, Konstanz, 2–4
September 2010.
Related Entries
Carnap, Rudolf | epistemology: Bayesian | evidence | Hempel, Carl |
induction: problem of | logic: inductive | probability, interpretations of |
statistics, philosophy of
Acknowledgments
I would like to thank Gustavo Cevolani, Paul Dicken, and Jan Sprenger for
useful comments on previous drafts of this entry, and Prof. Wonbae Choi
for helping me correcting a mistake.
Copyright © 2021 by the author

Vincenzo Crupi
68 Stanford Encyclopedia of Philosophy

Confirmation A4

Uploaded by

Copyright:

Available Formats

Confirmation A4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Confirmation A4

Uploaded by

Copyright:

Available Formats

pdf version of the entry

confirmation—has often been used more or less interchangeably with 2. Hypothetico-deductivism

2 Stanford Encyclopedia of Philosophy Spring 2021 Edition 3

4 Stanford Encyclopedia of Philosophy Spring 2021 Edition 5

6 Stanford Encyclopedia of Philosophy Spring 2021 Edition 7

8 Stanford Encyclopedia of Philosophy Spring 2021 Edition 9

10 Stanford Encyclopedia of Philosophy Spring 2021 Edition 11

12 Stanford Encyclopedia of Philosophy Spring 2021 Edition 13

14 Stanford Encyclopedia of Philosophy Spring 2021 Edition 15

simply because, if h ∧ k ⊨ e, then h ∧ q ∧ k ⊨ e , too, by the monotonicity

16 Stanford Encyclopedia of Philosophy Spring 2021 Edition 17

18 Stanford Encyclopedia of Philosophy Spring 2021 Edition 19

20 Stanford Encyclopedia of Philosophy Spring 2021 Edition 21

22 Stanford Encyclopedia of Philosophy Spring 2021 Edition 23

24 Stanford Encyclopedia of Philosophy Spring 2021 Edition 25

26 Stanford Encyclopedia of Philosophy Spring 2021 Edition 27

By (P0), (P1) and (P2), we thus have CP (h, e ∣ k) = f [P(h ∣ e ∧ k)],

28 Stanford Encyclopedia of Philosophy Spring 2021 Edition 29

Confirmation as firmness (F-confirmation, qualitative) CP (h, e1 ∧ e2 ∣ k) = CP (h, e1 ∣ k) + CP (h, e2 ∣ e1 ∧ k).

The point of qualitative F-confirmation is thus straightforward: h is said to

30 Stanford Encyclopedia of Philosophy Spring 2021 Edition 31

32 Stanford Encyclopedia of Philosophy Spring 2021 Edition 33

34 Stanford Encyclopedia of Philosophy Spring 2021 Edition 35

36 Stanford Encyclopedia of Philosophy Spring 2021 Edition 37

i. (P4) holds if and only if CP (h, e ∣ k) is a probability difference

38 Stanford Encyclopedia of Philosophy Spring 2021 Edition 39

does not hold that e entails h if and only if h entails e (h ⊨ e).

42 Stanford Encyclopedia of Philosophy Spring 2021 Edition 43

Bayesians that are unconvinced by the predictivist position are naturally

44 Stanford Encyclopedia of Philosophy Spring 2021 Edition 45

46 Stanford Encyclopedia of Philosophy Spring 2021 Edition 47

48 Stanford Encyclopedia of Philosophy Spring 2021 Edition 49

50 Stanford Encyclopedia of Philosophy Spring 2021 Edition 51

52 Stanford Encyclopedia of Philosophy Spring 2021 Edition 53

54 Stanford Encyclopedia of Philosophy Spring 2021 Edition 55

56 Stanford Encyclopedia of Philosophy Spring 2021 Edition 57

58 Stanford Encyclopedia of Philosophy Spring 2021 Edition 59

60 Stanford Encyclopedia of Philosophy Spring 2021 Edition 61

Continuum, pp. 428–449. Duhem’s Thesis Revisited”, Synthese, 177: 139–149.

62 Stanford Encyclopedia of Philosophy Spring 2021 Edition 63

64 Stanford Encyclopedia of Philosophy Spring 2021 Edition 65

Vassend, O.B., 2015, “Confirmation Measures and Sensitivity”, <https://plato.stanford.edu/archives/win2019/entries/scientific-

66 Stanford Encyclopedia of Philosophy Spring 2021 Edition 67

Other Internet Resources

Copyright © 2021 by the author

68 Stanford Encyclopedia of Philosophy

You might also like