ORIGINAL PAPER IN PHILOSOPHY OF SCIENCE

Carnap on concept determination: methodology
for philosophy of science

James Justus

Received: 28 October 2010 /Accepted: 26 May 2011 /Published online: 1 July 2011
# Springer Science+Business Media B.V. 2011

Abstract Recent criticisms of intuition from experimental philosophy and else-
where have helped undermine the authority of traditional conceptual analysis. As the
product of more empirically informed philosophical methodology, this result is
compelling and philosophically salutary. But the negative critiques rarely suggest a
positive alternative. In particular, a normative account of concept determination—
how concepts should be characterized—is strikingly absent from such work.
Carnap's underappreciated theory of explication provides such a theory. Analyses
of complex concepts in empirical sciences illustrates and supports this claim, and
counteracts the charge explication is only suitable for highly mathematical,
axiomatic contexts. Explication is also defended against the influential criticism it
is “philosophically unilluminating”.

1 Introduction

Recent criticisms of intuition from experimental philosophers and others have helped
undermine the authority of traditional conceptual analysis (see Cummins 1998;
Goldman and Pust 1998; Stich 1998; Hintikka 1999; Weinberg et al. 2001; Margolis

Euro Jnl Phil Sci (2012) 2:161–179
DOI 10.1007/s13194-011-0027-5

Thanks to John Carpenter, Mark Colyvan, Cory Juhl, Samir Okasha, Mark Sainsbury, Sahotra Sarkar,
Joshua Shepherd, and two anonymous reviewers for helpful comments. Audiences at the Sydney
University HPS Department and the third annual Sydney–Tilburg conference on philosophy of science
provided similarly valuable feedback. This work was generously supported by the Sydney Centre for the
Foundations of Science.

J. Justus (*)
Philosophy Department, Florida State University and University of Sydney, Tallahassee,
FL 32306, USA
e-mail: jjustus@fsu.edu
URL: http://philosophy.fsu.edu/content/view/full/35712

Keywords Explication . Carnap . Strawson . Definition . Methodology. Precision .

Ecological stability. Meaning . Concepts



and Laurence 2003; Nichols 2004; Bishop and Trout 2005; Ladyman and Ross
2007; Knobe and Nichols 2008; Weinberg et al. 2010). These criticisms show
intuitions are acutely sensitive to non-cognitive factors and often grossly mislead
about empirical phenomena (e.g. Griffiths et al. 2009), among other things. As the
product of more empirically informed philosophical methodology, this result is
compelling and philosophically salutary. But the negative critiques rarely suggest a
positive alternative. In particular, a normative account of concept determination—
how concepts should be characterized—is strikingly absent from such work. Most
cognition occurs in the currency of concepts so this deficiency goes to the core of
philosophy as well as science.

Carnap’s underappreciated theory of explication provides the alternative needed.
Section 2 describes explication and the epistemic rationale underlying it. One kind of
concept determination exemplifies the rationale: definition in science. Unlike lexical
definitions, scientific definitions readily sacrifice intuitive accord with the meaning
of the (often problematically vague) concept being defined to enhance formal rigor,
experimental testability, measurability in the field, mathematical and theoretical
unification, etc. The objective is improving the ability to describe, discover, and
understand the world by establishing relationships between nonscientific concepts
used to do so (or concepts from earlier stages of scientific development) and
concepts that better facilitate this goal. Given the remarkable epistemic success of
science, Carnap understandably thought concept determination in philosophy should
follow suit.

Rather than become a staple of philosophical methodology, explication has been
challenged on several grounds. For one, Carnap only casually indicated how a
handful of simple empirical concepts could be explicated. Explicating abstract
concepts in highly formal, axiomatic contexts (e.g. analyticity, mathematical truth,
probability) was the principal focus throughout his career. This invites the charge
that explication is not suitable for the complex empirical concepts and nonaxiomatic
contexts actually found in most scientific work and the grist for most philosophy of
science. Section 3 deflects this criticism by describing two applications of the
method to one such concept in biology, ecological stability. Besides showing how a
single concept can be explicated in distinct ways, the different explications also
illustrate priority relations between Carnap’s adequacy criteria for explication.

Conceptual flexibility is at the heart of explication: the license to deviate from the
intuitive meaning of concepts to reap epistemic rewards. Some claim this constitutes
a fatal flaw (e.g. Strawson 1963; Sorensen 1991; Eagle 2004; Loomis and Juhl
2006). Without strong similarity or even synonymy between the concept being
explicated and those providing the explication, P. F. Strawson (1963) indicted
explication as “utterly irrelevant” to the philosophical analysis of concepts. By
deviating from a concept’s prior meaning—no matter how problematic it may be—
one abandons the only legitimate source of philosophical insight.

Section 4 describes this criticism. Section 5 argues it caricatures explication and
that the view of conceptual analysis it presupposes involves a fundamental
misalignment of method and objective. Traditional conceptual analysis typically
targets concepts found in natural language that critical scrutiny reveals are
problematic. Intuitions are consulted to help clarify, rehabilitate, and systematize
relations between these concepts. Intuitions are therefore assumed to provide the best

162 Euro Jnl Phil Sci (2012) 2:161–179



guidance about concept determination, despite significant divergence among
philosophers (Cummins 1998) and across cultures (Weinberg et al. 2001).

If the goal of analysis is merely descriptive—to clarify and describe the current
meanings of concepts in natural language—this method is certainly relevant. But
gauged against this objective, intuitions are not the sole (and unlikely the best) guide
to conceptual content. Linguistic analyses of how language speakers actually employ
words associated with concepts (Jackson 1998), and studies of the mental
representation of concepts in cognitive science (Goldman and Pust 1998) also
provide such guidance, against which intuition must compete and be calibrated.

If the goal of analysis is normative—to improve our conceptual tools for
epistemic inquiry—recent criticisms (and much of the history of philosophy) show
that traditional conceptual analysis performs poorly. As an account of concept
determination steeped in scientific practice and grounded by its success, explication
is designed to achieve precisely this goal. Section 6 concludes that explication offers
philosophy, philosophy of science in particular, a cogent alternative methodology,
not because the scientific methodology on which it is modeled trumps intuition a
priori, but because of its vastly superior record.

2 Explication

Carnap’s most detailed exposition of explication (1950, Ch. 1) was intended as a
clarificatory preliminary before he explicated ‘degree of confirmation’ with a logical
relation between hypothesis and evidence statements. Explication is the transforma-
tion or replacement of an imprecise concept (explicandum) with a new concept(s)
(explicatum/explicata) subject to four adequacy criteria:

(i) Similarity (of meaning) to the explicandum;
(ii) Exactness1;
(iii) Fruitfulness; and,
(iv) Simplicity.

The first criterion was much weaker than what most philosophers at the time
required of a concept’s analysis and the philosophical definitions such analyses
yielded. Russell’s (1905) account of descriptions, for instance, required synonymy
between a proposition containing a description and the propositions characterizing it.
Moore (1942) also required synonymy between analyzed and analyzing concept(s).
But Carnap only required the explicatum be usable in most of the cases the
explicandum was; explication need not preserve meaning exactly. As the examples
discussed below show, even extensions were permitted to change. Carnap never
specified the precise degree of similarity needed, but the weak requirement of
preserving usage in most of the important cases was probably acceptable since
criterion (i) permitted “considerable differences” (1950, 7) in meaning. Explication
therefore threads between lexical definition, which requires identity of meaning in
principle, and stipulative definition, at least if stipulation need not preserve any
relationship between meanings (Belnap 1993; cf. Loomis and Juhl 2006, 287).

1 Following Carnap, ‘exactness’ and ‘precision’ are used interchangeably.

Euro Jnl Phil Sci (2012) 2:161–179 163



Explications are examples, Belnap (1993, 116) notes, of “perhaps the most
distinctively philosophical acts of definition,” in which, “one wants both to rely on
an old, existing meaning and to attach a new, proposed meaning.”

Why allow such latitude? Carnap had a compelling rationale: the increase in
fruitfulness afforded by sacrificing close similarity. This conceptual flexibility
accords, he emphasized, with scientific practice.2 Carnap illustrated with a simple
example. As the science of zoology emerged, the “prescientific” concept ‘Fish’,
which roughly meant ‘animal living in water’ and subsumed marine mammals,
evolved into a narrower concept ‘Piscis’ that excluded mammals. According to
Carnap (1950, 6), the original meaning was abandoned because:

zoologists found that the animal to which the concept Fish applies, that is,
those living in water, have by far not as many other properties in common as
the animals which live in water, are cold-blooded vertebrates, and have gills
throughout life. Hence the concept of Piscis defined by these latter properties
allows more general statements than any concept defined so as to be more
similar to Fish; and this is what makes the concept Piscis more fruitful.

From a contemporary perspective, this example is rather dated. Phylogeny, rather
than phenotypic similarity, is the primary reason mammals are excluded from Piscis
today (Ereshefsky 2001). The example nonetheless indicates that Carnap (1950, 7)
construed fruitfulness as facilitation of “general” statements. For nonlogical
concepts, this was gauged by how many law-like generalizations the explication
makes possible; for logical concepts, logical theorems.3 Explication therefore
violates the “conservativeness” or “noncreativeness” criterion commonly invoked
in theories of definition within formal languages: definitions should not extend the
set of theorems (other than those involving the defined concept) derivable within the
language (see Belnap 1993).

The exactness criterion requires “rules of [the explicatum’s] use…be given in an
exact form, so as to introduce the explicatum into a well-connected system of
scientific concepts” (1950, 7). Ideally, Carnap envisaged the system as a formal
language in which explicit meaning postulates and transformation rules specify
precise relations between terms that represent concepts, and in which sets of axioms
represent scientific theories, as described in the Logical Syntax of Language for
instance (Carnap [1934] (1937)). In such languages an explicatum could be
introduced by explicit definitions that specify relations between it and concepts
designated in the axioms, as well as theorems of different theories. This kind of
explication maximizes exactness, and as examples Carnap (1950, 16–17) cited
recursive definition of arithmetical operations with Peano’s axioms and Frege’s
definition of numerals (1963a, 935).

2 Some philosophically inclined scientists reached similar conclusions. In support of this flexibility, one of
the founders of ecology, Alfred Lotka (1925, Ch. 1), argued that concepts that have emerged for everyday
communication will likely not correspond to the concepts that best facilitate epistemic inquiry. Due to their
ambiguity or vagueness, concepts commonly assumed to describe the world in everyday communication,
such as ‘life’ or ‘causality,’ may fare poorly in this regard (Lotka 1925, 16).
3 Frege, Carnap’s instructor at Jena in the 1910s (Carnap 1963c), shared the view that definitions of logical
and mathematical concepts should be judged according to their fruitfulness. Definitions should “extend our
knowledge” by “drawing boundary lines that were not previously given at all” (Frege 1980, 100–101).

164 Euro Jnl Phil Sci (2012) 2:161–179



An explicit definition is one way to introduce an explicatum but there are others.
An explication may simply identify a term represented in the axioms as an
explicatum for a given explicandum. Satisfying the similarity criterion would then
depend on what the axioms say about the explicatum—as interpreted in a model of
the axioms—evaluated against the meaning of the explicandum. Fruitfulness could
be assessed in terms of the theorems the axioms entail.4 As such, the appeal of this
form of explication depends considerably upon the “implicit definition” the axioms
furnish for the explicandum. The implicit definition of the set concept furnished by
axiomatic set theory, for instance, arguably provides a highly fruitful explication (see
Muller 2004).

Carnap was well aware that such systematicity is rare among empirical rather than
formal sciences. Exactness for empirical concepts correspondingly focuses less on
maximizing precision via integration into axiomatized theories, and more on making
concepts more precise, particularly through partial formalization or quantification.
Carnap (1963a, 935–6) emphasized this explicitly when replying to Strawson
(1963): exactness only requires the explicatum be more precise than the
explicandum. He granted, in fact, that the explicatum may belong to a “more exact”
part of natural rather than formal or scientific language. This appropriately weakened
his previous claim that explication should integrate an explicatum into a well-
connected systems of scientific concepts (1950, 2), for two reasons. First, the more
exact part of natural language is presumably more exact because it contains more
precise nonscientific concepts. Second, Carnap elsewhere stressed that relations
between concepts in natural languages are generally opaque and imprecise compared
to conceptual relations in scientific languages (Carnap [1932] (1959)). With ample
room to enhance precision and to the extent increasing precision enhances
fruitfulness (see below), genuine explications can therefore be formulated in natural
language.

Carnap distinguished three types of non-logical concepts (in order of increasing
precision): classificatory, comparative, and quantitative. In normal scientific
progress, concepts typically evolve towards greater precision.5 ‘Temperature’ is an
example. Carnap (1950, 12) boldly hypothesized that in early, undocumented stages
in the development of natural language only classificatory concepts such as hot,
warm, and cold existed. Later, comparative concepts like warmer and colder
appeared. But not until science emerged was a quantitative concept of temperature
introduced to describe phenomena of heat and cold precisely. The evidentiary basis
for Carnap’s speculation appears scant, but it overlaps remarkably with what
meticulously detailed historical analysis reveals about the origin of the temperature
concept (see Chang 2004).

Differences between the quantitative temperature concept and the comparative
concepts of warmer and colder also show why an explicatum need not be useable in
all the cases the explicandum was. Organisms might experience an object to be

4 Note that since explication may be nonconservative, it may introduce inconsistencies into formal
systems.
5 Carnap carefully emphasized that the evolution towards greater precision was typical, not inevitable. For
example, he noted that adequate quantitative explicata for some psychological concepts may not exist, so
the focus should be finding fruitful comparative or classificatory explicata (Carnap 1950, 14). In these
cases, increasing precision can decrease fruitfulness.

Euro Jnl Phil Sci (2012) 2:161–179 165



warmer or colder at different times due to physiological changes even though the
object’s quantitative temperature remains constant. Such disparities do not indicate
temperature inadequately explicates ‘warmer’ or ‘colder’ because, according to
Carnap, “we have become accustomed to let the scientific concept overrule the
prescientific one in all cases of disagreement” (1950, 12–13). The scientific concept
has certainly supplanted the prescientific one in this case, but Carnap intended the
claim in full generality: if an explicandum and its explicatum differ in meaning,
gains in precision and fruitfulness afforded by the latter would justify it taking
priority over the former.6 This does not require, Carnap also stressed, that that the
explicandum must be eliminated.

But this analysis seems vulnerable to an obvious objection: isn’t Carnap merely
changing the subject? Rather than clarify the phenomenal concepts ‘warmer’ and
‘colder,’ Carnap has simply abandoned this goal for another, one of more scientific
interest. This worry underlies Strawson’s (1963) criticism that if conceptual
clarification is the objective, precision and fruitfulnesss are irrelevant if meaning is
not closely preserved. In fact, he argued these desiderata usually thwart rather than
advance the goal of “philosophically illuminating” the explicandum because they
typically encourage departure from the original, often imprecise meaning. Before
describing the contrasting philosophical methodology underlying this criticism in
Sections 4, Section 3 deflects a different kind of objection: that explication may be
well-suited for highly theoretical concepts—such as those actually explicated in
great detail and technical sophistication by Carnap (e.g. analyticity, degree of
confirmation, semantic information [Carnap [1934] (1937), 1947, 1950; Carnap and
Bar-Hillel 1953])—but that it is ill-equipped for the complex concepts actually
employed in empirical science. The seemingly oversimplified explications of rather
rudimentary empirical concepts that Carnap presents unfortunately encourages this
view. Many scientific examples could be marshaled (e.g. Carnap 1977 on entropy,
Chang 2004 on temperature), but two very different explications of the complex
concept of ecological stability show the criticism is mistaken particularly clearly.
They also illustrate how explications can enhance exactness and fruitfulness in
distinct ways.

3 Precision, fruitfulness, and ecological stability

The idea there is a “balance of nature” was a staple of the schools of natural
philosophy from which biology emerged, long before the term ‘ecology’ was even
coined (Egerton 1973). Some early ecologists continued this tradition by attempting
to derive the existence of a “natural balance” in biological populations from
organismic metaphors and anthologies with physical systems (see Kingsland 1995).
Not until the second half of the twentieth century was the concept of a balance of
nature rigorously characterized, and the predominantly metaphysical speculations

6 Simplicity (criterion [iv]) was subsidiary to the other adequacy criteria for Carnap. He noted that many
concepts of well-confirmed and well-developed scientific theories are quite complicated, and suggested
simplicity only matters when deciding between explicata of comparable precision, fruitfulness, and
similarity to the explicandum. The value of simplicity in explication is therefore secondary and context-
dependent, similar to its value in other aspects of science such as inference and explanation (Sober 1988).

166 Euro Jnl Phil Sci (2012) 2:161–179



about its cause superseded with scientific hypotheses about its basis. Although
significant uncertainty and controversy still remains about what features of an
ecosystem’s dynamics should be considered its stability (see Justus 2008a), most
ecologists adopt one of two kinds of accounts of ecological stability.

The first appeals to dynamic systems theory and is frequently found in
mathematical modeling in ecology (e.g. May 1974; Logofet 1993; Kot 2001). Let
a vector x(t) (t designates time) represent an ecological system, such as a biological
community, each vector component representing a distinct part of the system, such as
a species.7 Let x* represent the system at equilibrium. The system at equilibrium is
Lyapunov stable if and only if:

8" > 0ð Þ 9d > 0ð Þ x t0ð Þ � x»
�
�
� < d ) 8t � t0ð Þ xðtÞ � x»

�
�
� < "

�
�
�

� ��
�
�

� �

; ð1Þ

where ε and δ are real numbers, x(t0) represents the system at some initial time t0,
and ‘|·|’ designates a Euclidean distance metric. Informally, (1) says x* is Lyapunov
stable if systems beginning in a neighborhood of x* remain near it after being
perturbed to the state x(t0) at t0, i.e. that the effects of perturbations are
circumscribed.

The second characterization of ecological stability is more frequently found in
recent empirical studies in ecology (e.g. Tilman 1999; Lehman and Tilman 2000).
Let Bi be a random variable designating the biomass of species i in an n-species
community C assayed during some time period, and let Bi designate the expected
value of Bi for the time period it is assayed. Tilman (1999) defined the temporal
stability of C, St(C), as:

StðCÞ ¼
Pn

i¼1
Bi

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Pn

i¼1
Var Bið Þ þ

Pn

i¼1; j¼1
i6¼j

Cov Bi; Bj
� �

v
u
u
t

; ð2Þ

where Var designates variance and Cov designates covariance. Unlike (1), St
characterizes ecological stability in terms of the variability of biomasses of species in
the community. The idea motivating this account is that if two series of abundances
are plotted across time, the more stable one exhibits less fluctuation (Lehman and
Tilman 2000).

Both accounts satisfy the exactness criterion by formulating precise explicata, but
in utterly different ways. The first increases precision by proposing that the formal
statement (1) characterizes ecological stability. The second, however, increases
precision by characterizing ecological stability as a quantitative property whose
value is determined by the mathematical function (2).

(1) and (2) increase precision in different ways, and they are fruitful in different
ways. The first explication integrates the concept of ecological stability into a well-
developed mathematical theory that provides analytic methods for evaluating the

7 Specifically, the components of x(t) are variables that usually represent densities, abundances, or
biomasses of species in the community. Components of the equilibrium state x* therefore represent these
densities, abundances, or biomasses at equilibrium.

Euro Jnl Phil Sci (2012) 2:161–179 167



stability of mathematical models of biological communities. This increase in formal
rigor and tractability, however, has nontrivial costs (see Justus 2008b). Accurate
mathematical models describing the dynamics of real-world ecological systems are
usually unavailable. Without such models, evaluation of whether (1) holds must be
done with time series data on x(t), which is extremely difficult. Strictly speaking, for
instance, the universal quantification in the consequent of (1) [∀t≥t0] requires
monitoring system behavior over indefinite time periods. Whether (1) holds also
depends on system behavior after one perturbation (represented by the displacement
from x* to x(t0)). Empirical evaluation of this behavior in the field is often obscured
when environments subsequently (and frequently) perturb systems.

In fact, the infeasibility of empirically evaluating (1) was a disadvantage Lehman
and Tilman (2000, 535) cited in favor of (2): temporal stability is “readily observable
in nature” because evaluating (2) only requires empirical measurement of biomasses,
rather than the difficult empirical assessment of dynamic system behavior (1)
requires. Once the concept is made empirically tractable in this way, the role stability
may have in system dynamics and its possible correlations and causal relationships
with other features of biological communities can be rigorously investigated (see
Tilman 1999). Greater measurability with (2), however, also has nontrivial costs.
Unlike Lyapunov stability, temporal stability has no theoretical underpinning in a
mathematical theory of system dynamics, such as dynamics systems theory
(Hinrichsen and Pritchard 2005). (2) also does not characterize ecological stability
in terms of how communities respond to perturbation, which seems indispensable to
the concept (see Justus 2008a, §5), and therefore threatens to violate the similarity
criterion. Ideally, an explication would enhance fruitfulness while retaining sufficient
similarity to the explicandum, without tradeoffs. The tradeoffs in fruitfulness
between (1) and (2)—in this case between a more empirically and a more
theoretically oriented conception of fruitfulness—indicate this may be rare in
science.

(1) and (2) explicate ecological stability in different, nonequivalent ways.
Lyapunov stability, for instance, is a concept whose application requires that a
community initially be at equilibrium, whereas temporal stability places no
restrictions on a community’s initial state. The existence of nonequivalent character-
izations of the same concept would typically be taken to indicate an underlying
problematic ambiguity of that concept. But developing multiple, distinct explicata
for even an unambiguous explicandum is permissible and sometimes advisable on
Carnap’s account of explication. Since preservation of strong similarity of meaning
is not required, different explicata can differ from the meaning of the explicandum in
distinct ways. Pursuing several distinct explications for a given concept may be
useful since different explicata can be fruitful in distinct ways, similar to the merits
of different logics (e.g. Carnap, R. [1934] (1937)), and the different merits of (1) and
(2). This permissive view of the relationship between the meaning of the explicatum
and explicandum, however, prompted the most serious criticism of explication.

Before describing that criticism, it is worth considering more generally why
explication is designed to reflect the indispensable role of precision in scientific
methodology. Explication requires increasing precision—at one point Carnap even
says it is “the only essential requirement” (1963a, 936)—but its value does not
derive from precision alone. Rather, precision is paramount because it usually

168 Euro Jnl Phil Sci (2012) 2:161–179



enhances fruitfulness. Vague concepts are rarely, if ever, components of the well-
confirmed generalizations the fruitfulness criterion targets. The problematically
vague concept ‘cause’, for instance, is not a technical term of contemporary science
and currently plays no role in any reputable scientific theory (Dowe 2000).8

Concepts like ‘germ’ or ‘life’ arguably share this status, as ‘ecological stability’ did
before being explicated.

Making concepts more precise, in contrast, often facilitates in developing and
discovering well-confirmed generalizations. ‘Fish’, ‘temperature’, and the much
more complicated and precise characterizations of ‘energy’ that emerged in physical
mechanics and electrodynamics are paradigmatic examples. Without sufficiently
precise concepts, it is difficult if not impossible to derive predictions from statements
containing them. Without such predictions, in turn, statements cannot be confirmed
or disconfirmed. Testing predictions of hypotheses and theories against data is a
staple of scientific methodology that has clearly proved to be an epistemically
reliable basis for inference. Precision facilitates this methodology and its success
therefore grounds the exactness criterion.

Although increasing precision usually enhances fruitfulness, there are exceptions.
Two cases of precision increases not producing greater fruitfulness can be
distinguished. First, the current state of scientific knowledge may be the problem.
At a given time, scientists may simply fail to recognize the law-like statements more
precise explicata would make possible. Greater precision may not enhance
fruitfulness in a stronger sense if the concept being explicated represents something
essentially imprecise. Carnap thought some psychological concepts might be
examples. Beliefs states, for example, may be best represented with qualitative
rather than fully quantitative probabilities (Walley 1991, Ch. 5). In such cases greater
precision may not increase, and even decrease fruitfulness due to facts about the
world rather than just limitations of scientists’ epistemic acumen (see Justus 2006).

Serious doubts about whether law-like generalizations exist in sciences such as
biology and psychology also suggest the notion of fruitfulness should be broadened
to recognize the improvement an explicatum can have in discovering well-confirmed
generalizations through enhancing experimental testability, measurability in the field,
theoretical unification, mathematical rigor, etc. rather than just in the production of
law-like generalizations. Explications of ecological stability illustrate this more
expansive and defensible notion of fruitfulness.

4 Strawson’s criticism

Explication contrasts sharply with the method of concept determination predomi-
nantly employed in philosophy. Call it ‘conceptual analysis.’ As aptly represented by
Strawson (1963, 514–518), conceptual analysis assumes there are underlying

8 The qualifiers ‘contemporary’ and ‘currently’ are intended to grant that the concept could have such a
role in future science. Among extant analyses of causation, Salman (1997) and Dowe’s (2000) account of
causal interactions as intersections of world-lines that involve exchange of conserved quantities is one
candidate; the causation concept sometimes found in studies of Bayes nets is another (see Spirtes et al.
2001). These accounts offer characterizations of the concept precise enough to make possible, for instance,
its empirical detection and even empirical measurement of causal interaction strength.

Euro Jnl Phil Sci (2012) 2:161–179 169



principles governing concepts and their use in ordinary communication, or that would in
highly idealized, maximally rational, and philosophically informed versions of this
communication. Elucidating these principles is the goal of conceptual analysis, and
philosophical definition as its end product.9 Details about the way concepts are actually
deployed, intuitions about them, and thought experiments that probe their conditions
and range of application provide the raw material from which the elucidation is
derived. Strawson hoped that judicious scrutiny of these factors would uncover the
“logical behavior which ordinary concepts exhibit” (Strawson 1963, 513) which the
“elusive, deceptive modes of functioning of unformalized linguistic expressions”
(1963, 512) often mask. Almost three decades later, Strawson offered an analogy to
help clarify the methodology (1992, 7):

Just as we may have a working mastery of the grammar of our native language,
so we have a working mastery of [our] conceptual equipment. We know how to
handle it, how to use it in thought and speech. But just as the practical mastery
of the grammar in no way entails the ability to state systematically what the
rules are which we effortlessly observe, so the practical mastery of our
conceptual equipment in no way entails the possession of a clear, explicit
understanding of the principles which govern our handling of it, of the theory
of our practice. So—to conclude the analogy—just as the grammarian, and
especially the model modern grammarian, labours to produce a systematic
account of the structure of rules which we effortlessly observe in speaking
grammatically, so the philosopher labours to produce a systematic account of
the general conceptual structure of which our daily practice shows us to have a
tacit and unconscious mastery.

Finding such principles would clarify the underlying “structure” of our concepts.
As such, Strawson claimed the method provides the best basis for defensible
definitions of concepts and the best way, perhaps only way, to resolve the conceptual
confusions and the philosophical problems they generate.

Contrasted with this methodology, Strawson objected that explication is not
similarly “philosophically illuminative”:

however much or little the constructionist technique [i.e. explication] is the
right means of getting an idea into shape for use in the formal or empirical
sciences, it seems prima facie evident that to offer formal explanations of key
terms of scientific theories to one who seeks philosophical illumination of
essential concepts of non-scientific discourse, is to do something utterly
irrelevant—is a sheer misunderstanding, like offering a text-book on
physiology to someone who says (with a sigh) that he wished he understood
the workings of the human heart. (Strawson 1963, 504–505)

Part of his basis for this claim was the observation that:

it seems in general evident that the concepts used in non-scientific kinds of
discourse could not literally be replaced by scientific concepts serving just the

9 “Paradigmatic conceptual analyses offer definitions of concepts that are to be tested against potential
counterexamples that are identified via thought experiments” (Margolis and Laurence 2008, §2.1).

170 Euro Jnl Phil Sci (2012) 2:161–179



same purposes; that the language of science could not in this way supplant the
language of the drawing-room, the kitchen, the law courts and the novel.
(Strawson 1963, 505)

Since scientific concepts cannot replace nonscientific ones and retain the same
functionality, there is no reason to expect a “well-connected system” of the latter
would shed any light on the former. It follows, Strawson asserted, “that typical
philosophical problems about the concepts used in non-scientific discourse cannot be
solved by laying down the rules of exact and fruitful concepts in science. To do this
last is not to solve the typical philosophical problem, but to change the subject”
(1963, 506). Instead of yielding insights about concepts that generate philosophical
puzzles, explication pursues the project of finding precise substitutes for them. It
thereby abandons the only subject matter Strawson believed could resolve the
puzzles. Desertion does not constitute illumination.

Strawson located the motivation for explication in an untenable, even dogmatic
dismissal of traditional philosophical problems (Strawson 1963, 517). Conceptual
analysis, in contrast, involves careful analysis of the actual concepts and usage that
generate the problems. Thus, for Strawson explicata will only provide insights about
concepts (explicanda) if there is a clear account of the relations between them. But
this requires an accurate description of the explicanda, as conceptual analysis at least
might provide. One could only think otherwise, Strawson (1963, 513) suggests, “if
one is led away from the purpose of achieving philosophical understanding by the
fascination of other purposes, such as that of getting on with science.”

5 Philosophical methods and objectives

Carnap and Strawson’s disagreement primarily concerns the similarity criterion.
Carnap was willing to sacrifice close similarity to achieve greater precision and
fruitfulness. For Strawson, as one deviates from a concept’s content one abandons
the goal of philosophically illuminating it in favor of something else.

5.1 Common ground

Despite their differences, there is a significant point of potential accord between
Carnap and Strawson unappreciated by contemporary commentators (e.g. Loomis
and Juhl 2006; Maher 2007). Carnap and Strawson could agree about one aspect of
the issue: understood as an empirical method of inquiry about concepts, conceptual
analysis can help clarify the current meanings of explicanda. Appropriating
Strawson’s (1992) analogy, just as linguists analyze linguistic practices of natural
language speakers to determine syntax and semantics, how philosophers think about
concepts helps identify their meanings and elucidate their interrelationships. How
terms that designate concepts are employed in everyday communication also helps
pinpoint these meanings and relationships (Jackson 1998),10 and studies of the

10 This assumes that usage patterns for terms of natural language provides reliable information about (but
does not necessarily determine) meaning.

Euro Jnl Phil Sci (2012) 2:161–179 171



mental representation of concepts in cognitive science are similarly salient (Goldman
and Pust 1998). Cognitive studies and linguistic analyses of term use are therefore
relevant to conceptual analysis understood in this empirical sense. In fact, Carnap
(1963a, 920) explicitly advocated this kind of approach:

It seemed rather plausible to me from the beginning that there should be an
empirical criterion for the concept of the meaning of a word or a phrase, in
view of the fact that linguists traditionally determine empirically the
meanings, meaning differences, and shifts of meanings of words, and that
with respect to these determinations they reach a measure of agreement
among themselves … In [Carnap 1955], I tried to show the possibility of
giving operational rules for testing hypotheses concerning the intensions of
predicates of a natural language, on the basis of responses by the users of this
language.

‘Intension’ is Carnap’s term for a predicate’s meaning. Developing this method
for determining predicate intensions in natural languages (Carnap 1955) was part of
Carnap’s response to Quine’s criticisms of an analytic-synthetic distinction. If
intensions can be identified with this method, the analyticity of statements
concerning these intensions can be assessed. The significant congruence in Carnap
and Strawson’s own response to Quine is striking given their disparate philosophical
commitments (see Grice and Strawson 1956).

In fact, Carnap accords an important role to this type of empirical conceptual
analysis in explication. Before proposing an explicatum, Carnap stressed that an
explicandum’s meaning should be clarified as much as possible through
informal methods (Carnap 1950, 3–5). This preliminary clarification helps
determine whether an explicandum is ambiguous, which suggests multiple
explicata are appropriate, and whether an explicandum is too vague to even
explicate.11 If an explicandum is too vague, presumably its similarity to potential
explicata cannot be evaluated, and attempts to explicate it would therefore be
pointless.12 Carnap (1963a, 936) suggested, furthermore, that such informal
clarification alone may resolve some philosophical confusions, particularly those
that emerge when natural language terms are used outside their normal scope in
everyday communication. Experimental philosophers’ emphasis on empirically
uncovering the (sometimes indefensible) factors responsible for intuitions and
other conceptual judgments, among the proverbial “folk” as well as the
philosophically trained, makes an important contribution to this element of
explication.

Carnap could even grant that proverbial “armchair” conceptual analysis—
where the predominant focus is what intuitions reveal—can help identify the
current meaning of an explicandum. Philosophers are (usually) competent natural
language speakers and especially keen to subtle differences in meaning. Their
understanding of concepts at least constitutes relevant empirical data about their

11 “[Philosophers] often immediately start to look for an answer [to questions like ‘What is X?’ where X is
some concept] without first examining the tacit assumption that the terms of the question are at least
practically clear enough to serve as the basis for an investigation, for an analysis or explication” (Carnap
1950, 4).
12 Carnap (1950) did not make this argument explicitly.

172 Euro Jnl Phil Sci (2012) 2:161–179



meaning (Goldman and Pust 1998). Thus, far from “dismiss problems about
ordinary language” (Maher 2007, 333), Carnap recognized a role for traditional
philosophical approaches to such problems in the first step of explication. This fact
is too often overlooked by critics (Strawson 1963; Eagle 2004; Loomis and Juhl
2006) and advocates (Maher 2007) of explication alike. Of course, since
philosophers’ intuitions about concepts likely reflect their theoretical commit-
ments, their intuitions will likely disagree (because their commitments typically
diverge), and fail to reflect those of the general populace (because these
commitments are not shared) (Dowe 2004).

5.2 Method–objective mismatch

Despite this potential accord, “clarification of the explicandum” was merely
preparatory for Carnap. Although attempting to clarify an explicandum’s
meaning with the methods considered above can be a valuable pre-explication
preliminary, explication requires a very different methodology, one that
sanctions tradeoffs between preserving meaning and enhancing precision and
fruitfulness.

Aligning methodological resources with philosophical objectives is the central
issue. As Section 4 indicates, Strawson utilized a more substantive conception of
conceptual analysis than the empirical version Carnap could accept. On this view,
arguably the dominant view in philosophy, conceptual analysis is not simply
empirical inquiry into what philosophically trained individuals intuit about
concepts.13 The results of such inquiries, and definitions based on them, would
seem to lack any normativity. They would merely describe how a fairly small,
generally unrepresentative group of individuals construe certain concepts, not
prescribe how they should be.

Rather than accept this attenuated view, intuitions are taken to confer special
status on such analyses: “[conceptual analysis] attempts to show the natural
foundations of our logical, conceptual apparatus, in the way things happen in the
world, and in our own natures” (Strawson 1963, 516). Its purpose is therefore to
uncover the “foundation of our concepts in natural facts, and to envisage alternative
possibilities, it is not enough to have a sharp eye for linguistic actualities” (1963,
517). Intuitions about this “natural foundation” supposedly supply conceptual
analyses with prescriptive force. Deviation from intuition to increase precision or
fruitfulness, regardless of how epistemically advantageous and inductively
warranted, is therefore indefensible.

The fundamental question confronting this view of conceptual analysis is what, if
anything, do intuitions demonstrate beyond how a set of humans understand and use

13 One potential exception is Jackson (1998, 44), who is “suspicious” about the view that conceptual
analysis provides insights about the world. He calls this “immodest” conceptual analysis, and expresses
doubts about his earlier use of it in the knowledge argument (Jackson 1986). Jackson (1998) defends
instead a “modest” conceptual analysis that attempts to determine the folk theory he assumes underlies
concept use in natural language and on which cognitive and linguistic studies bear. But if this is the goal, it
is unclear what legitimate role modest conceptual analysis has in metaphysics, as Jackson believes it has.
As traditionally understood, metaphysics is concerned with uncovering the fundamental nature of the
world, not anthropological theorizing (cf. Jackson 1998, Ch. 3).

Euro Jnl Phil Sci (2012) 2:161–179 173



particular concepts?14 Evaluating this question requires distinguishing two objec-
tives. One is whether intuitions can help clarify the current content of concepts,
which there is little reason to doubt (as discussed above). The stronger, more
problematic goal is that intuition provides reliable insights about the “foundation of
our concepts in natural facts,” insights that potentially compete with scientific
claims, and which indicate how concepts should be determined. The first goal is
descriptive, the second is normative. Given its poor track record on the second (e.g.
Harman 1994; Hintikka 1999; Griffiths et al. 2009), the onus is on intuition
proponents to defend its epistemic utility and role in concept determination.
Different attempts to do so have been made, but several compelling criticisms cast
serious doubt about its feasibility.

A review of the literature underlying the growing recognition that intuition is
epistemically suspect cannot be undertaken here. It has been thoroughly elsewhere
(e.g. Levin 2004; Ladyman and Ross 2007; Knobe and Nichols 2008). But without
retracing well-worn steps, it is worth briefly recounting why it is unlikely a
convincing defense will be forthcoming. What is required, Goldman and Pust (1998)
note, is a convincing explanation of how intuitions, which are mental events of some
kind, are a reliable source of information about extra-mental facts. With visual
perception, for instance, this is supplied by a complex set of causal relationships that
includes the objects being perceived, the light reflected from them, the perceiver’s
retinas, and the visual cortex. Currently, some details of these relationships and those
involved in other types of perception may be unknown, but it is clear what kind of
connection exists between the extra-mental objects or properties perceived and our
perceptions of them. The same cannot be said for intuition. The common contention
that concepts are abstracta and truths about them are grasped through a non-
empirical faculty of intuition obscures more than it clarifies (see Levin 2004). The
apparently significant diversity of intuitions across cultures only exacerbates this and
other difficulties (Stich 1998; Weinberg et al. 2001).

6 Conclusion: conceptual methodology for philosophy of science

If recent criticisms of intuition are sound, much of philosophy is in need of a new
methodology. This does not mean traditional conceptual analysis has no philosophical
future. If description is the objective—clarifying the current meanings of terms,
elucidating the putative ‘folk’ theories underlying concepts, etc.—conceptual analysis
is one of many possible sources of information (anthropological, cognitive, and
linguistic studies are others). And for some philosophical issues, acquiring descriptive

14 Another difficulty specific to Strawson’s grammar analogy will not be pursued here. Strawson assumed
an underlying “system” or “structure” exists that governs our concepts and explains how the ordinary
language terms referring to them facilitate communication (1963, 513; 1992, Ch. 2). The form of this
conceptual system was never made clear. Strawson called it “logical” (1963, 513), however, and suggested
conceptual analysis should reveal logical relations between concepts, i.e. necessary and sufficient
conditions. Whether there are conceptual relations of this kind is an open question and psychological
experiments seem to disconfirm Strawson’s claim (e.g. Rosch and Mervis 1975). Based on evidence that
context-sensitive prototypes govern concept use, for instance, Ramsey (1998) argues that the traditional
goal of conceptual analysis of finding necessary and sufficient conditions for concepts that are resistant to
intuitive counterexamples is misguided.

174 Euro Jnl Phil Sci (2012) 2:161–179



information about concepts may yield sufficient clarity to resolve disagreements.
Ethical judgments, for instance, often concern whether an evaluative concept applies
to an act or decision. Intuitions about hypothetical scenarios may help reveal
unrecognized features of evaluative concepts relevant to these judgments. But relative
to this descriptive goal, intuition-based conceptual analysis contributes but one voice.
It must compete against results from experimental philosophy and non-philosophical
disciplines, and its results may be (and likely are) less credible about conceptual
content. The sensitivity of intuition to seemingly non-cognitive factors such as race,
sex, socioeconomic class, and others uncovered by experimental philosophers further
challenges the legitimacy of traditional conceptual analysis, even for descriptive
projects of this kind, and encourages their emphasis on developing an empirically
informed alternative methodology.

This descriptive characterization of conceptual analysis obviously fails to reflect
its normative ambitions in other areas, such as epistemology, metaphysics, and
philosophy of science. But if the goal is providing normative guidance about
concepts by enhancing their cognitive credentials, the critique of intuitions has
sweeping implications. If, specifically, the objective is aligning our concepts with a
conceptual system that better describes and better facilitates in discovering the
structure of the world, then the absence of a credible epistemic basis for intuition
entails it is not normatively controlling. Given this objective, explications need not
preserve intuitive meaning. Requiring similarity ensures some continuity in the
epistemically driven conceptual changes explication counsels. But the same rationale
that sanctions sacrificing strong similarity may also advise abandoning rather than
attempting to explicate some concepts. As a method of conceptual engineering,
explication accords no privilege to the conceptual apparatus we inherit and which
provides the fodder for traditional conceptual analysis.

Freed from intuitive constraint, concepts and their analysis become just another
set of epistemic tools evaluated according to the function they, and the broader
cognitive and linguistic structure in which they participate, are intended to serve.
Carnap (1963b, 938–939) championed this pragmatic view:

In my view, a language, whether natural or artificial, is an instrument that may
be replaced or modified according to our needs, like any other instrument … A
natural language is like a crude, primitive pocketknife, very useful for a
hundred different purposes. But for certain specific purposes, special tools are
more efficient, e.g., chisels, cutting-machines, and finally the microtome. If we
find that the pocket knife is too crude for a given purpose and creates defective
products, we shall try to discover the cause of the failure, and then either use
the knife more skillfully, or replace it for this special purpose by a more
suitable tool, or even invent a new one. [Strawson’s] thesis is like saying that
by using a special tool we evade the problem of correctly using the cruder tool.
But would anyone criticize the bacteriologist for using a microtome, and assert
that he is evading the problem of correctly using a pocketknife?

In reply, Strawson (1963, 505) correctly emphasized that precise concepts such as
those found in science will likely never replace most of the imprecise concepts used
in everyday communication (see §4). Loomis and Juhl’s (2006, 291) response to

Euro Jnl Phil Sci (2012) 2:161–179 175



Carnap’s analogy is similarly accurate: “pocketknives are not replaceable by
microtomes for most ordinary uses” (emphasis added). But concepts being sufficient
for ordinary uses and providing effective means for everyday communication does
not entail they are the best or even passable conceptual tools for epistemic inquiry. If
this is the ‘specific purpose’ to which Carnap alludes, transforming concepts into
more precise and fruitful ones is warranted, even if departing from their current
content is unavoidable. With respect to this normative goal, it is thoroughly opaque
what conceptual analysis contributes.

Loomis and Juhl (2006, 291) raise another, related criticism: “someone who was
having trouble using a pocket-knife in an ordinary circumstance would not be helped
in the least by being shown the workings of a microtome” (emphasis added). Although
this may be true, its critical force against explication and Carnap’s instrumental view
of language is unclear. Continuing the analogy, if the pocketknife corresponds to a
term of natural language denoting a concept and difficulty arises in an ordinary
circumstance—i.e. one in which the term is normally used—then it seems the troubled
concept user does not understand the term’s actual meaning and should consult a
lexicon. If the problem persists, it is likely the individual is expecting more of the
concept than it delivers in ordinary circumstances where it functions unproblemati-
cally. The microtome corresponding to the more precise explicatum was, after all, only
recommended as a “special tool” invoked for “certain specific purposes.”

A putative sharp distinction between philosophical and scientific concepts, and a
belief their analysis requires disparate evaluative standards underlies these criticisms.
On this basis, Strawson could agree with Carnap about the advantages of precision for
properly scientific concepts such as ‘ecological stability,’ but insist that exclusively
philosophical concepts such as ‘knowledge’ or ‘personal identity’ require a different
approach. Explication’s role in enhancing the relation between the former and well-
confirmed generalizations is clear, the objection might continue, but it is not for the latter.

There are several responses to this objection:

(1) Many (perhaps most) concepts that philosophers of science study seem to fall
on the “properly scientific” side of the ledger, including ‘entropy,’ ‘field,’
‘fitness,’ ‘force,’ ‘information,’ ‘natural selection,’ ‘signal,’ and ‘species’
among others. If explication is apt for clearly scientific concepts, it is apt for
much of the work in philosophy of science. Intuitions can play a role in helping
clarify these explicanda, just as intuitions about thought experiments had a
crucial role in the theorizing that led Einstein to relativity theory (Norton 1991).
But accord with intuition does not itself confer epistemic merit on explicata.

(2) Besides Carnap’s influential explicatory analyses, much highly regarded work
in the philosophy of science mirrors his approach. The deft interplay of
empirical and theoretical work displayed in Skryms’ (2010) cogent study of
meaning with sender-receiver signaling models is a particularly clear recent
example, but numerous others share this approach (e.g. Crupi et al. 2007).15

15 Recall that the degree of precision appropriate for an explication depends on what a concept represents
(see §3). Some concepts studied by philosophers of science, like some psychological concepts, may
concern essentially imprecise phenomena. If scientific explanation is indelibly intertwined with an
amorphous sense of subjective understanding, highly precise explicata for it may be less fruitful than less
precise ones (see Trout 2002).

176 Euro Jnl Phil Sci (2012) 2:161–179



(3) Many concepts central to philosophy of science seem to defy label as strictly
philosophical or scientific and it is generally unclear on what basis, besides
mere disciplinary division, such a distinction between putative types of
concepts could be drawn. There are numerous examples: ‘confirmation,’
‘disposition,’ ‘inference,’ ‘law,’ ‘model,’ ‘natural kind,’ ‘observation,’ ‘proba-
bility,’ ‘representation,’ ‘space,’ and ‘time’ among others. Since nothing about
these concepts suggests intuition is any more reliable regarding their analysis,
explication remains the only viable method of concept determination.

Admittedly, the ultimate goal of explication—finding well-confirmed
generalizations—may seem orthogonal to the purpose of analysis in philosophy
of science. But this conflates distance with irrelevance. Analyses by
philosophers of science and scientists often differ in degree of abstraction and
generality, but they do not differ in kind or in their ultimate goal. The
immediate aim of analyses of concepts in philosophy of science is improving
the understanding of their role (or potential role) in scientific practice, as well
as the bearing empirical and theoretical results have on characterizations of
these concepts and statements containing them. The epistemic success of
science (and failure of intuition) merits this focus and analyses of concepts
often found in the ‘Theory’ or ‘Concept’ sections of scientific publications
share precisely this objective. Although sometimes quite removed from
conceptual work, the ultimate payoff in understanding philosophical work
affords is measured in the same enduring currency scientists use to measure
success: well-confirmed generalizations. Philosophers of science and scientists’
interests and tools typically differ in level of generality and degree of
abstraction, but they toil for the same fruit:

He who wishes to investigate the questions of the logic of science must,
therefore, renounce the proud claims of a philosophy that sits enthroned above
the special sciences and must realize that he is working in exactly the same
field as the scientific specialist, only with a somewhat different emphasis.
(Carnap [1934] (1937), 332)

Rather than “offer nothing” to answer Strawson’s objection (Maher 2007, 333),
Carnap sought to undermine the standard philosophical methodology in favor of a
more scientifically cognizant, inductively warranted alternative. The compelling
criticisms of intuition more than a half century later reveal the prescience of Carnap’s
efforts.

References

Belnap, N. (1993). On rigorous definitions. Philosophical Studies, 72, 115–146.
Bishop, M., & Trout, J. D. (2005). The pathologies of standard analytic philosophy. Noûs, 39, 696–714.
Carnap, R. ([1932], 1959). The elimination of metaphysics through logical analysis of language. In A. J.

Ayer (Ed.), Logical positivism (pp. 60–81). Glencoe: Free.
Carnap, R. ([1934], 1937). The logical syntax of language. London: Routledge.
Carnap, R. (1947). Meaning and necessity. Chicago: University of Chicago Press.
Carnap, R. (1950). Logical foundations of probability. Chicago: University of Chicago Press.

Euro Jnl Phil Sci (2012) 2:161–179 177



Carnap, R. (1955). Meaning and synonymy in natural languages. Philosophical Studies, 7, 33–47.
Carnap, R. (1963a). P. F. Strawson on linguistic naturalismk. In P. Schilpp (Ed.), The philosophy of Rudolf

Carnap (pp. 933–940). Lasalle: Open Court.
Carnap, R. (1963b). W. V. Quine on logical truth. In P. Schilpp (Ed.), The philosophy of Rudolf Carnap

(pp. 915–921). Lasalle: Open Court.
Carnap, R. (1963c). Intellectual autobiography. In P. Schilpp (Ed.), The philosophy of Rudolf Carnap (pp.

3–84). Lasalle: Open Court.
Carnap, R. (1977). In A. Shimony (Ed.), Two essays on entropy. Berkeley: University of California Press.
Carnap, R., & Bar-Hillel, Y. (1953), An outline of a theory of semantic information. Technical Report No.

247: Research Laboratory of Electronics, MIT.
Chang, H. (2004). Inventing temperature: Measurement and scientific progress. New York: Oxford

University Press.
Crupi, V., Tentori, K., & Gonzalez, M. (2007). On Bayesian measures of evidential support: theoretical and

empirical issues. Philosophy of Science, 74, 229–252.
Cummins, R. (1998). Reflections on reflective equilibrium. In M. R. DePaul & W. Ramsey (Eds.),

Rethinking intuition: The psychology of intuition and its role in philosophical inquiry (pp. 113–128).
New York: Rowman and Littlefield.

Dowe, P. (2000). Physical causation. New York: Cambridge University Press.
Dowe, P. (2004). Causation and misconnections. Philosophy of Science, 71, 926–931.
Eagle, A. (2004). Twenty one arguments against propensity analyses of probability. Erkenntnis, 60, 371–

416.
Egerton, F. (1973). Changing concepts of the balance of nature. The Quarterly Review of Biology, 48,

322–350.
Ereshefsky, M. (2001). The poverty of the Linnaean Hierarchy: A philosophical study of biological

taxonomy. Cambridge: Cambridge University Press.
Frege, G. (1980). Foundations of arithmetic. Evanston: Northwestern University Press.
Goldman, A., & Pust, J. (1998). Philosophical theory and intuitional evidence. In M. R. DePaul & W.

Ramsey (Eds.), Rethinking intuition: The psychology of intuition and its role in philosophical inquiry
(pp. 179–197). New York: Rowman and Littlefield.

Grice, H. and Strawson, S. (1956). “In Defense of a Dogma.” Philosophical Review, 65, 141–158.
Griffiths, P., Machery, E., & Linquist, S. (2009). The vernacular concept of innateness. Mind and Machine,

24, 605–630.
Harman, G. (1994). Doubts about conceptual analysis. In M. Michael & J. O’Leary-Hawthorne (Eds.),

Philosophy in mind (pp. 43–48). Dordrecht: Kluwer.
Hinrichsen, D., & Pritchard, A. (2005). Mathematical systems theory I: Modelling, state space analysis,

stability and robustness. New York: Springer.
Hintikka, J. (1999). The emperor’s new intuitions. Journal of Philosophy, 96, 127–147.
Jackson, F. (1986). What Mary didn’t know. Journal of Philosophy, 83, 291–295.
Jackson, F. (1998). From metaphysics to ethics: A defense of conceptual analysis. New York: Clarendon.
Justus, J. (2006). Loop analysis and qualitative modeling: Merits and limitations. Biology and Philosophy,

21, 647–666.
Justus, J. (2008a). Complexity, Diversity, Stability. In S. Sarkar & A. Plutynski (Eds.), A companion to the

philosophy of biology (pp. 321–350). Malden: Blackwell.
Justus, J. (2008b). Ecological and Lyapunov stability. Philosophy of Science, 75, 421–436.
Kingsland, S. (1995). Modeling nature: Episodes in the history of population ecology (2nd ed.). Chicago:

University of Chicago Press.
Knobe, J., & Nichols, S. (Eds.). (2008). Experimental philosophy. New York: Oxford University Press.
Kot, M. (2001). Elements of mathematical ecology. Cambridge: Cambridge University Press.
Ladyman, J., & Ross, D. (2007). Everything must go: Metaphysics naturalized. New York: Oxford

University Press.
Lehman, C. L., & Tilman, D. (2000). Biodiversity, stability, and productivity in competitive communities.

The American Naturalist, 156, 534–552.
Levin, J. (2004). The evidential status of philosophical intuition. Philosophical Studies, 121, 193–

224.
Logofet, D. (1993). Matrices and graphs: Stability problems in mathematical ecology. Ann Arbor: CRC.
Loomis, E., & Juhl, C. (2006). Explication. In S. Sarkar & J. Pfeifer (Eds.), The philosophy of science: An

introduction (pp. 287–294). New York: Routledge.
Lotka, A. J. (1925). Elements of mathematical biology. New York: Dover.
Maher, P. (2007). Explication defended. Studia Logica, 86, 331–341.

178 Euro Jnl Phil Sci (2012) 2:161–179



Margolis, E., & Laurence, S. (2003). Should we trust our intuitions?: Deflationary accounts of the analytic
data. Proceedings of the Aristotelian Society, 103, 299–323.

Margolis, E., & Laurence, S. (2008). Concepts. In E. N. Zalta (Ed.), Stanford encyclopedia of philosophy
(Fall 2008 edition) <http://plato.stanford.edu/entries/concepts>.

Moore, G. E. (1942). A reply to my critics: Analysis. In P. Schilpp (Ed.), The philosophy of G. E. Moore
(pp. 660–667). Lasalle: Open Court.

Muller, F. A. (2004). The implicit definition of the set-concept. Synthese, 138, 417–451.
Nichols, S. (2004). Folk concepts and intuitions: From philosophy to cognitive science. Trends in

Cognitive Sciences, 8, 514–518.
Norton, J. D. (1991). Thought experiments in Einstein’s work. In T. Horowitz & G. Massey (Eds.),

Thought experiments in science and philosophy (pp. 129–148). Lanham: Rowman & Littlefield.
Ramsey, W. (1998). Prototypes and conceptual analysis. In M. R. DePaul & W. Ramsey (Eds.), Rethinking

intuition: The psychology of intuition and its role in philosophical inquiry (pp. 161–178). New York:
Rowman and Littlefield.

Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of categories.
Cognitive Psychology, 8, 382–439.

Russell, B. (1905). On denoting. Mind, 14, 479–493.
Salman, W. (1997). Causality and explanation: A reply to two critiques. Philosophy of Science, 64, 461–

477.
Skyrms, B. (2010). Signals. New York: Oxford University Press.
Sober, E. (1988). Reconstructing the past: Parsimony, evolution and inference. Cambridge: MIT.
Sorensen, R. (1991). Vagueness and the desiderata for definition. In J. H. Fetzer, D. Shatz, & G. N.

Schlesinger (Eds.), Definitions and definability: Philosophical perspectives (pp. 71–110). Boston:
Kluwer.

Spirtes, P., Glymour, C., & Scheines, R. (2001). Causation, prediction, and search (2nd ed.). Cambridge:
MIT.

Stich, S. (1998). Reflective equilibrium, analytic epistemology and the problem of cognitive diversity. In
M. R. DePaul & W. Ramsey (Eds.), Rethinking intuition: The psychology of intuition and its role in
philosophical inquiry (pp. 95–112). New York: Rowman and Littlefield.

Strawson, P. F. (1963). Carnap’s views on constructed systems versus natural languages in analytic
philosophy. In P. Schilpp (Ed.), The philosophy of Rudolf Carnap (pp. 503–518). Lasalle: Open Court.

Strawson, P. F. (1992). Analysis and metaphysics: An introduction to philosophy. New York: Oxford
University Press.

Tilman, D. (1999). The ecological consequences of biodiversity: A search for general principles. Ecology,
80, 1455–1474.

Trout, J. D. (2002). Scientific explanation and the sense of understanding. Philosophy of Science, 69, 212–
233.

Walley, P. (1991). Statistical reasoning with imprecise probabilities. New York: Chapman and Hall.
Weinberg, J., Nichols, S., & Stich, S. (2001). Normativity and epistemic intuitions. Philosophical Topics,

29, 429–460.
Weinberg, J., Gonnerman, C., Buckner, C., & Alexander, J. (2010). Are philosophers expert intuiters?

Philosophical Psychology, 23, 331–355.

Euro Jnl Phil Sci (2012) 2:161–179 179

http://plato.stanford.edu/entries/concepts

	Carnap on concept determination: methodology for philosophy of science
	Abstract
	Introduction
	Explication
	Precision, fruitfulness, and ecological stability
	Strawson’s criticism
	Philosophical methods and objectives
	Common ground
	Method–objective mismatch

	Conclusion: conceptual methodology for philosophy of science
	References