1

The Nature of Inquiry

Aristotle famously wrote that philosophy begins in wonder.

Philosophical inquiry is, of course, prompted by much more than
this. Curiosity, puzzlement, annoyance, distress, and even deep
anger can prompt us to it. Although these states are highly diverse,
they have this much in common: they induce us to raise questions
that we want our inquiry to resolve.
In philosophy, the central questions concern such topics as the
nature of reality, the character and extent of our cognitive powers,
and the content and status of practical norms. These questions
include, but are certainly not limited to, the following:

Why is there something rather than nothing?

Does God exist?
Do we have free will?
How is the mind related to the body?
Is there such a thing as human nature?

Philosophical Methodology: From Data to Theory. John Bengson, Terence Cuneo, and Russ Shafer-­Landau,
Oxford University Press. © John Bengson, Terence Cuneo, and Russ Shafer-­Landau 2022.
DOI: 10.1093/oso/9780192862464.003.0002
 The Nature of Inquiry 11

How is knowledge of the external world possible, if it is?

How do thought and language refer to the world, if they do?
What is the nature of meaning?
Are there objective normative standards?
Is there a single, fundamental principle of morality, and if so, what is it?
What is it to live well?

Naturally, philosophers also wrestle with a welter of more circum-

scribed questions, such as:

Is existence a property?
Which abilities are required for omnipotence?
Does free will imply the ability to do otherwise?
What is belief?
Is gender a biological category?
What is the difference between justified true belief and knowledge?
Is it possible to refer to abstract objects?
Is meaning normative?
What is it for a standard to be objective?
What distinguishes moral principles from merely prudential ones?
Does living well require being virtuous?

To engage in philosophical inquiry is to endeavor to address ques-

tions such as these, and those yet more fine-­grained. This means
treating such questions as open, aiming to respond to them in as
satisfying a manner as possible. Ideally, the culmination will be
the successful resolution of those questions, the closing of inquiry.
Our focus is on philosophical inquiry in its theoretical mode.
Such inquiry is a species of a genus—theoretical inquiry. As such,
it will help us understand our target by gaining a clearer view of
12 The Nature of Inquiry

the genus itself. We’ll do this by tackling two fundamental

­questions about such inquiry. The first concerns its structure: what
are the basic elements of, or stages in, this enterprise? While
sometimes a messy business in practice, we identify two common
threads in diverse instances of theoretical inquiry. Our second
question focuses on the goals of such inquiry: what is its point or
purpose? We distinguish different types of goals, arguing that
­successful theoretical inquiry terminates in a certain kind of
understanding.

1. The Structure of Inquiry

Inquiry comes in a variety of forms, each involving a process or
activity aimed at some end or goal. In practical inquiry, we engage
in deliberation aimed at the performance of an action. In the­or­et­
ic­al inquiry, we engage in reflection aimed at the provision of a
theory. We’ll examine the goals of theoretical inquiry in detail in
the next section. In this section, we analyze its structure.1
In our view, theoretical inquiry is best modeled as a process
structured around two stages. We call the first the data collection
stage, since it centers on the gathering of considerations (the data)
that must somehow be handled when addressing a set of ques-
tions with which an inquiry is concerned. We call the second the
theorizing stage, since it is designed to address those questions by
developing theories that handle the data.

1
Some maintain that inquiry is best analyzed in terms of a distinctive attitude or
aim (see, e.g., Friedman 2017, 302–3 and Smith 2020, 184). We remain neutral on this
issue, providing an account not of inquiry itself but rather of its structure (its components
and their organization). That said, our account of this structure might be parlayed into
a novel analysis of inquiry.
 The Structure of Inquiry 13

Each stage has three main elements. Both data collection and
theorizing have inputs, outputs, and a procedure or method, which
takes one from the former (input) to the latter (output).2 The two
stages are intimately connected insofar as the outputs of the first
stage are the inputs to the second. Together, when all goes well,
they constitute the transition from the opening of inquiry—­
concerning a set of questions—to its closing. Inquiry concludes
with the provision of a theory, a set of claims that inquirers advance
in the course of satisfying the method’s criteria in the second stage.

Opening Closing

Question

Question Data
Theorizing Theory
Collection
Question
Stage 1 Stage 2
Procedure yields Method takes Realizes goal
data as output. data as input.

The Structure of Theoretical Inquiry

We’ll examine each stage of inquiry in detail in subsequent

chapters. Here we note simply that this two-­stage model is intended
to be a theory of the basic structure of theoretical inquiry, as distinct
from a description of such inquiry as it occurs on the ground. The
differences between the two are real: the latter is a temporally
extended process that agents carry out more or less competently,
while the former is not. Though these differences exist, the struc-
ture of theoretical inquiry is hardly divorced from actual inquiries.
The activities in which inquirers engage approximate the structure

2
To regiment terminology, we’ll be speaking of ‘procedures’ in the case of data
collection, and ‘methods’ in the case of theorizing.
14 The Nature of Inquiry

of theoretical inquiry (or its elements) to different degrees and can

be evaluated by the extent to which they conform to this struc-
ture. While our model will find a home for the many legitimate
things that inquirers actually do, it is not our intention to fully
characterize all the workings and difficulties of actual instances
of theoretical inquiry, but rather to identify its basic components
and to specify their configuration.3 (Hereafter, as in the title of this
chapter, we will often speak simply of ‘inquiry,’ eliding ‘the­or­et­
ic­al’ for the sake of brevity.)
To illustrate our two-­stage model with an example drawn from
philosophy, consider a group of inquirers focused on the question
of whether God exists. Stage one represents their efforts to iden-
tify considerations that must somehow be taken into account by a
theory that adequately answers this question. Such data include:
that the world exists, that many natural objects are highly complex
and organized, that the world contains a tremendous amount of
suffering, that numerous individuals have sincerely claimed to
experience God, that a great many people report having failed to
do so even after making sincere attempts, and so on. Stage two
represents the efforts of these inquirers—theists and non-­theists
alike—to develop theories that account for these and other data.
If their theories are good enough, this would thereby close the
question that opened the inquiry.

3
Our model allows that inquirers may inquire without a fully determinate question
in mind, or in pursuit of an open-­ended question such as “What is such and such like?”
Inquiries can also overlap with, nest in, and compose other inquiries. Indeed, thinkers
often undertake multiple inquiries at once, which combine to form a single, evolving
inquiry. For example, investigations regarding defeat, context, and stakes may target
questions uncovered in the course of investigating knowledge and justification; all
of these inquiries (and others) may be components of one and the same inquiry
into skepticism.
 The Structure of Inquiry 15

Philosophers sometimes flirt with models of theoretical inquiry

that do not mark any of its stages. For example, on Robert
Stalnaker’s familiar and influential picture of “the abstract struc-
ture of inquiry,” inquirers are treated as believers seeking to nar-
row the space of possible worlds to a certain subset, one containing
those and only those worlds that answer the queries that prompted
inquiry.4 This picture looks seriously incomplete. For it does not
tell us what it is for inquiry to unfold; it merely observes that it
does. When analyzing the structure of inquiry, what is wanted is
insight into the components of inquiry, and how those compo-
nents are related so as to constitute a certain kind of process.
Whatever its merits, the possible worlds picture of inquiry does
not provide such insight.
The solution isn’t simply to posit multiple stages. Consider, for
example, a five-­stage view of inquiry advanced by John Dewey,
which comprises

(i) a felt difficulty, (ii) its location and definition, (iii) suggestion of a possible
solution, (iv) development by reasoning of the bearing of the suggestions,
(v) further observation and experiment leading to its acceptance or rejection.5

But a process culminating in observation and experiment is

insufficient to capture inquiries regarding a wide range of theo-
retical questions. We aren’t going to solve the Liar paradox, figure
out why there is something rather than nothing, or ascertain the
­distribution of primes through observation and experiment—­
especially if these are empirical undertakings, as Dewey insists.

4
Stalnaker (1984, ix). We’ll waive worries about how this picture could account for
beliefs regarding specific necessary and essential truths.
5
Dewey (1910, 72; cp. 1938, Ch. VI).
16 The Nature of Inquiry

A different sort of problem afflicts the model of inquiry made

famous in Descartes’ unpublished treatise, Rules for the Direction of
the Mind. Consider, in particular, Rules 4 and 5:

We need a method if we are to investigate the truth of things.

The whole method consists entirely in the ordering and arranging of the
objects on which we must concentrate our mind’s eye if we are to discover
some truth. We shall be following this method exactly if we first reduce com-
plicated and obscure propositions step by step to simpler ones, and then,
starting with the intuition of the simplest ones of all, try to ascend through
the same steps to a knowledge of the rest.6

This model posits a structure with two main stages. However, this
structure is of the wrong sort, being narrowly focused on execut-
ing the Cartesian project of rendering clear and distinct a series of
individual ideas or propositions. While this is perhaps one type of
theoretical inquiry, it is plainly not the only game in town. Inquiries
across the sciences and humanities do not drill down to what is
“simplest” and then, using intuition, “ascend” from there to “a
knowledge of the rest.” A satisfactory account of the structure of
inquiry should be perfectly general, applying to every token of
the type.
More promising than the possible worlds, Deweyan, and
Cartesian approaches is a dialectical model, according to which
inquiry consists in formulating and clarifying candidate answers

6
Descartes (1628/1985; emphases added). Later, in the Discourse on Method, Descartes
identifies four components to his favored method: evidence (assent to nothing that can
be doubted), division (divide a problem into parts), order (begin with the simplest
parts), and exhaustion (omit no parts). The worries we raise about Descartes’ model
also apply to models confined to the discovery of necessary or essential truths (as in the
Socratic Picture of inquiry described by Bronstein 2016, Ch. 8).
 The Structure of Inquiry 17

to the questions that open inquiry, and then developing and

assessing arguments for and against those answers.7 This model
posits a structure while seemingly avoiding the parochialism of
the previous two models. But this alternative does not fairly repre-
sent how inquiry unfolds, or what it aims to uncover. When
inquiring about a murder, Sherlock Holmes does not simply list
suspects—candidate answers to the question “Who did it?”—and
then assemble arguments supporting their guilt or innocence.
Rather, as the detective makes clear in A Study in Scarlet, his inquiry
begins with the collection of data:

“You don’t seem to give much thought to the matter in hand,” I [Watson] said
at last, interrupting Holmes’ musical disquisition.

“No data yet,” he answered. “It is a capital mistake to theorize before you have
[them].”8

Only after collecting data does the detective pursue a view of the
murder that resolves his query. Such a view is more than just the
proposition that a certain suspect is guilty. Holmes might achieve
a knockdown argument whose conclusion is that Professor
Moriarty committed the crime in the drawing room with a vial
of poison, but still have no clue why Moriarty was moved to
homicide, how he did it with that particular substance, or the
extent to which this crime fits with Moriarty’s past behavior. In
these ways, a dialectical treatment of Holmes’ criminal inquiry

7
Different versions of the dialectical model may introduce important details, such
as the role of disputation, questioning, or juxtaposition of contradictory answers. See,
e.g., Adler (1927, vi), Bird (1953), Gadamer (1960/1989), Mueller (1965), Rosen (1982),
Hintikka (1999), and Rescher (2006).
8
Doyle (1887, Pt. I, Ch. III). In The Adventure of the Copper Beeches, Holmes exclaims:
“Data! Data! Data! . . . I can’t make bricks without clay.”
18 The Nature of Inquiry

Illustration 1 Sherlock Holmes collecting data.

omits im­port­ant dimensions of his venture. Likewise, a dia­lect­ic­al

treatment of theoretical inquiry misses or obscures the role
of data, which are fundamental to inquiries in a broad range of
fields—biology, economics, and philosophy of religion (recall
our example above), to name a few. The dialectical model also
distorts the project of constructing a theory that handles the
data. The latter goes beyond the mere provision of argument: as
we noted in the Introduction, the conclusion of an argument
does not a theory make. So, although the dialectical model
describes an undertaking that may play a role in theoretical
inquiry, it is certainly not the whole story about that process.
 The Structure of Inquiry 19

This brings us back to the model we’re proposing. There are

­several reasons to prefer a model that incorporates the two stages
of inquiry (data collection and theorizing) that we’ve distinguished.9
First, doing so furnishes a substantive precisification of the idea
that inquiry consists in a process of reflection aimed at the provi-
sion of a theory. Our model cracks open this black box, identify-
ing not only the starting point and the endpoint, but also two
stages in the process that takes inquirers from the former to the
latter.
Second, it does this while avoiding parochialism. Our model
identifies two stages that do not themselves enforce a commit-
ment to anything like the Cartesian project or Deweyan experi-
mentalism. Nor do the stages render either verboten. Rather, our
model represents steps by which inquirers of any persuasion, and
into any topic, transition from the opening of a question to the
grasp of a theory that resolves it.
Third, a model that highlights data and theorizing is well posi-
tioned to account for the depth and breadth of inquiry. Both fea-
tures were on display in our discussion of dialectic, which noted
that inquiry may incorporate, but is not exhausted by, the provi-
sion of arguments. This is just what we should expect if inquiry
consists in constructing a theory (theorizing) that handles a
potentially wide range of considerations (data).
Fourth, our model makes sense of a success condition on the­
or­ies noted in the Introduction: a theory is successful only when
it handles the data, its inputs. It also explains the Holmesian

9
In a pair of rich but neglected essays on philosophical methodology, Collingwood
(1933, Ch. VIII) and Castañeda (1980, passim) also insist on the importance of data
collection and theorizing. However, neither presents them as the two basic stages of
inquiry, and their conception of each is markedly different from ours.
20 The Nature of Inquiry

injunction against theorizing in the absence of any data.

Additionally, the model sheds light on various aspects of partic-
ular real-­world inquiries, such as the ways in which they can be
incomplete (as when they omit data collection, exclude the­or­iz­
ing, or make only partial contributions to either) or deficient (as
when they do at least one of these things poorly). In keeping with
these remarks, we’re not intent on gate-­keeping: incomplete or
deficient inquiry still counts as inquiry.
Fifth, by distinguishing the roles played by data collection and
theorizing, our model makes it possible to examine the basic
elem­ents of inquiry and their relations in particular cases. Even
though many actual instances of inquiry only imperfectly realize
the two stages, attending to them helps to sharpen questions
about how, in practice, to get from the opening of inquiry to its
closing. In other words, modeling inquiry in terms of these two
stages is not just theoretically interesting but also practically
important. For it enables us to scrutinize what it takes to do each,
and to do each well—an important first step toward improving our
efforts at inquiry.
To appreciate how distinguishing the two stages does this,
notice that it helps us to pinpoint various methodological com-
mitments, which can then be subject to proper assessment. In
phil­oso­phy, the longstanding debate between metaethical realists
and their expressivist rivals can serve as an illustration.
Realist views affirm the existence of a moral reality that is ob­ject­
ive, in the sense that it is not of our own devising in the way that
(say) the norms of etiquette are. And they hold that moral thought
and discourse sometimes genuinely represent this reality.
Expressivists typically deny at least one of these claims. Divergent
methodological commitments lie at the heart of this dis­agree­ment.
 The Structure of Inquiry 21

Each view makes assumptions about data. Realists typically

assume that the metaethical data consist of commonsensical or
intuitive claims, some of which appear to imply that there is an
objective moral reality that moral thought and discourse represent.
Expressivists ordinarily assume that the data primarily concern
moral language and moral judgment, emphasizing connections
among moral discourse, moral thought, and action.
Likewise, each view makes assumptions about theorizing.
Realists typically hold that a satisfactory theory ought to endorse
the implications of common sense or intuition, abandoning
them only when challenges to them cannot be adequately met.
Expressivists, for their part, typically assume that a satisfactory
theory must be parsimonious, not positing any features that are
explanatorily dispensable. They add that a satisfactory theory
must be sufficiently continuous with what the natural sciences tell
us about the world.
These assumptions are not idle; rather, they have fueled much
of the debate between realists and expressivists. Adherents of each
view draw very different conclusions largely because of their
divergent assumptions about data and theorizing. Realists typ­ic­
al­ly conclude that since their theory endorses the implications of
common sense or intuition, and has adequately met the relevant
challenges, theirs is the view to beat.10 Expressivists ordinarily
conclude that their view alone handles the data about language

10
See, e.g., Nagel (1986, 143–4), McNaughton (1988, 40–1), Brink (1989, Chs. 1–2),
Shafer-Landau (2003, Chs. 1–3), and Enoch (2019, 30–3). While not a realist, Harman
(2010) promotes a general conservative methodology that squares with the approach
favored by these authors.
22 The Nature of Inquiry

and judgment in a way that is sufficiently parsimonious and

nat­ur­al­is­tic, and therefore earns top marks.11
Realists and expressivists often fail to make their commitments
regarding data and theorizing explicit, and rarely pause to defend
them even when they do. Nor is metaethical debate unique in
these respects. To identify just two other examples: the debates
between realists and nominalists about universals exhibit a similar
dynamic, as do those between defenders and critics of the a prio-
ri.12 The former members of each pair frequently begin with an
unarticulated commitment to intuitive data and conservative the­
or­iz­ing, while the second regularly assume that it is safe to prefer a
theory that is both parsimonious and naturalistic.
Here we take a different approach. Though our remarks thus far
have ranged over all kinds of theoretical inquiry, our aim will be to
explicitly state and defend a series of methodological commit-
ments about data and theorizing that, in our view, anyone engaged
in philosophical inquiry should accept. We’ll articulate these com-
mitments over the coming chapters, offering detailed accounts of
the nature of data, procedures for data collection, and the process
of theorizing. Along the way we’ll indicate how our commitments
relate to those of pragmatists, conceptual analysts, experimental
philosophers, proponents of reflective equilibrium, fans of mod­el­
ing, and various others. As will become evident, the commitments
we advocate are both substantive and diverge in important ways
from those that frequently guide philosophical inquiry. At the
same time, we do not discard current practices, but rather gather

11
See, e.g., Blackburn (1993, Part II and 1998, Ch. 3) and Gibbard (2003, Chs. 1–2).
Cp. Ayer (1936, Ch. 6), who adds a commitment to a now-­defunct theory of meaning.
12
See, e.g., Bealer (1993), Dorr (2007), and Carmichael (2010) on universals, and the
exchange between Bonjour (2014) and Devitt (2014) on the a priori.
 The Goals of Inquiry 23

and organize the bits of methodological wisdom they contain.

We’ll argue that when philosophers adhere to the commitments
we present, the theories they develop are poised to realize the epi­
stem­ic goals of inquiry. It is to these goals that we now turn.

2. The Goals of Inquiry

Aristotle’s Posterior Analytics is naturally read as proposing that an
ultimate proper goal of theoretical inquiry is the acquisition of
epistêmê—where this is not mere ordinary knowledge, but a special
kind of epistemic ideal that many commentators designate with
the term ‘understanding.’13 We endorse a version of this proposal.
Let us begin by explaining what we have in mind by an ‘ultimate
proper goal’ of theoretical inquiry. We allow that the set of pos­
sible goals of inquiry is open-­ended. It may include mere psycho-
logical states, such as what Charles Sanders Peirce referred to as
“the fixation of belief,”14 or accomplishments such as happiness or
salvation. No doubt it includes positive epistemic states such as
justified belief, knowledge, and understanding.
We take some such goals to be ‘proper’ because, while there
might be no such thing as the correct goal of a given activity, some
goals are more fitting to an activity than others. This is particularly
true of goals whose realization constitutes a type of success or
excellence at that activity. As an example, consider investigative
journalism, which has a variety of proper goals that include
uncovering official malfeasance, identifying social inequalities, or

13
See, e.g., Burnyeat (1981, 102–3 and 112), who describes Aristotle’s proposal as
the “vision of Plato’s Republic transferred to the individual sciences.”
14
Peirce (1877, §4).
24 The Nature of Inquiry

exposing reparable inefficiencies within an organization. Such

journalism is done well when it succeeds at these aims. By con-
trast, humiliating personal enemies or publicizing boring arcana
are improper journalistic goals, for their achievement does not
imply a job well done. When the activity is theoretical inquiry,
undertaken for its own sake, success or excellence does not consist
in scoring points, winning accolades, or cultivating good habits.
Rather, it resides in making an intellectual improvement, in the
form of an epistemic or alethic achievement. In our view, this is
the têlos of inquiry; it is what inquiry is for.
We call a proper goal ‘ultimate’ just in case its realization is suf-
ficient to resolve inquiry in a fully successful way, reaching the
point at which, to put it simply, there is no more work to be done.
The notion of the fully successful resolution of inquiry is not
purely descriptive, but has a normative component as well, which
admits of various explications. According to one plausible expli-
cation, such resolution renders it appropriate for the inquirers’
curiosity, wonder, or puzzlement to be relieved thereby.15 Though
we wish to remain neutral between this and other candidate expli-
cations, we affirm that any plausible explication will demand not
merely partial and piecemeal insights, but rather comprehensive
and systematic illumination.
Not all proper goals are ultimate. For example, to be conscien-
tious at each stage of one’s investigation may be a proper goal of
inquiry, since it represents a type of success or excellence at this
activity. But conscientious inquiry may produce an erroneous pic-
ture of its target, and so does not qualify as an ultimate proper

15
Pritchard (2016, 34). We are wary of non-­normative explications, such as those
that emphasize the first-­personal experience of a “Eureka moment” (cp. Trout 2002).
 The Goals of Inquiry 25

goal. Even if inquirers were to correct these errors, their job would
hardly be over, since true belief can be unjustified and still leave
open many of the most significant questions about the target of
inquiry.16 Were inquirers to attain a true belief that was also justi-
fied, there would remain more work to do. For such a belief, or
even a set of such beliefs, does not by itself guarantee the sort of
comprehensive and systematic illumination required to resolve
inquiry in a fully successful way (hereafter, simply ‘resolve’ inquiry).
The point becomes clear when we consider that even know­ledge
fails to secure the guarantee. For such an achievement might be
gained on the basis of testimony or rote memorization, rendering
it compatible with obliviousness to the considerations that sup-
port or explain the answer. Ordinary knowledge is also congruent
with messes or tangles, as when new information does not resolve
cognitive dissonance, but instead induces it. In either case, inquiry
isn’t finished.
Importantly, knowledge also may incorporate confusion and
muddle of various sorts. A grade-­school class might know that π is

16
We thereby reject Dubois’ (1898, 16) contention that the aim of inquiry is “simple
truth,” a position echoed by a number of contemporary philosophers (cp. Niinuluoto
1984; Haack 1993; Anderson 1995). Other critics of this view include Kitcher (1990
and 1993), Bright (2017), and Nado (2019), as well as proponents of the simple
know­ledge view discussed in the text (e.g., Williamson 2000; Bird 2007; Whitcomb
2010; Mirrachi 2015; Beebee 2018; Sosa 2019; Kelp 2021). There are a clutch of alterna-
tive conceptions of the aim of philosophical inquiry in particular. For example, Gutting
(2009 and 2016, 325) privileges “knowledge of distinctions and of the strengths and
weaknesses of various pictures and their theoretical formulations”; Nozick (1981,
21–2) and Wilson (2014, 145–50 and 2017, 92) celebrate the amassing of theoretical
options; Deleuze and Guattari (1996, 5) hold that “the object of philosophy is to create
concepts that are always new”; Adorno (1993, 102) views philosophy as “an effort to
express things one cannot speak about”; and many philosophers insist that securing
agreement is part of philosophy’s aim. The considerations that follow apply mutatis
mutandis to these proposals; we raise doubts about the need for agreement in §3.
26 The Nature of Inquiry

the ratio of a circle’s circumference to its diameter, while being

confused about the concept ratio. For instance, they could mis­
taken­ly conceive of a ratio as (say) something whose decimal
expansion must terminate or repeat. In such a case, they would
know the answer to a key question about π—“Is π the ratio of a
circle’s circumference to its diameter?”—without having reached
the point at which further inquiry regarding this question would
be otiose. Philosophers who know the answer to a central question
in the philosophy of race—“Is race a social construct?”—could be
in a similar position. Whatever the correct answer, if they are
confused about the concept social construct, mis­taken­ly
conceiving of it as (say) requiring explicit group agreement, their
inquiry regarding this central question isn’t closed. For there
would be cracks in their grasp of the answer that make it unfitting
to call it a day; their knowledge of the answer would not appropri-
ately relieve their curiosity.
The core problem is not solved by piling on additional
­know­ledge—for instance, knowledge of what a ratio or social con-
struct is. For any such knowledge could itself incorporate conceptual
confusions that render it defective. It follows that thinkers could
have wide-­ranging knowledge of the answers to the questions
prompting their inquiries, but fail to resolve them. While ordinary
knowledge of an answer is a proper goal, it falls short of resolving
inquiry, and thus being an ultimate proper goal in our sense.17
Inquirers don’t need to pursue such a goal—proper goals are
good, too! Still, the differences between the various goals are

17
This line of reasoning entails the rejection of two norms of inquiry proposed by
Friedman (2017 and 2019): the “Don’t Believe and Inquire” norm and the “Ignorance
Norm,” according to which one ought not inquire, if one believes or knows, re­spect­
ive­ly. Cp. Archer (2018, §5).
 Theoretical Understanding 27

s­ ignificant, given three purposes that they serve. Such goals

furnish reference points for selecting promising questions to pursue
in the course of inquiry, as these questions are ones whose
reso­lution will secure those goals; they provide a yardstick for
assessing the merits of theories, which are criticizable to the extent
that they do not deliver the goods that such goals specify; they also
function as the basis for evaluating the methods used to construct
those the­or­ies.
For example, a method of theorizing might yield true belief and
nothing more. But such a method is inadequate if what we are
seeking is an ultimate proper goal of inquiry. If the outputs of the­
or­iz­ing should be assessed, as we believe, by reference to whether
they enable realization of such a goal, then when evaluating extant
philosophical methods (Chapter 4) and developing and defending
our own preferred method (Chapters 5 and 6), we’ll want to have
squarely in view the question of whether they manage this feat. It
will be helpful to have a term for a method that does. Accordingly,
we propose to call a method ‘sound’ just in case satisfaction of its
criteria thereby positions inquirers to achieve an ultimate proper
goal of inquiry. If a method is sound, then (ceteris paribus) in­quir-
ers have strong reason to endorse its outputs.
Our own view, as noted, is that theoretical understanding is an
ultimate proper goal. We turn now to articulating the central char-
acteristics of the relevant type of understanding, before arguing
for its appeal as such a goal.

3. Theoretical Understanding
We can reveal the central characteristics of understanding by con-
sidering some of the properties that a theory of a given domain
28 The Nature of Inquiry

must possess in order to provide understanding of that domain to

thinkers who fully grasp the theory.18 (We’re working under the
assumption that a domain, or subject matter, consists of objects,
properties, and the like, and is individuated by a set of questions.
We’ll speak sometimes of understanding a domain and at other
times of understanding with respect to a set of questions.19) Given
the number of controversies surrounding the nature of under-
standing, we wish to make clear which features we take this epi­
stem­ic achievement to exemplify. Theoretical understanding, as
we’ll construe it, is the state that agents possess just when they
fully grasp a theory with the following six properties.
First, the theory possesses a high degree of accuracy, since largely
inaccurate theories will fail to dispel confusion (a characteristic of
misunderstanding).
Second, the theory is reason-­based, in the sense that it is posi-
tively supported by considerations, beyond mere coherence, that
speak in favor of its accuracy. For in the absence of such support,
signing on to the theory would be arbitrary or haphazard (again, a
characteristic of misunderstanding).
Third, the theory is robust, answering a multitude of questions
about the most important features of the domain under investiga-
tion. A theory that neglects or dodges such questions leaves out
just what’s needed to yield comprehension.

18
We speak of ‘fully’ grasping not to imply perfection in every respect, but rather to
draw a contrast with a merely partial grasp.
19
Like inquiries, domains can overlap with, nest in, and compose one another. They
can also be cognized independently of fully successful inquiry into the set of questions
that individuate them. While there is much more to say about subject matters, to our
knowledge nothing in our project hangs on how they are further theorized.
 Theoretical Understanding 29

Fourth, the theory is illuminating, in that its answers must at least

sometimes be not just general but also genuinely explanatory,
going beyond a mere description of those features to explain why
each exists or is instantiated.
Fifth, the theory is orderly, not simply offering such feature-­
specific explanations but also affording a broader view of the
domain by revealing how those (and other) features, as well as the
proposed explanations, gel or hang together—for example, by
exposing basic relations or systematic connections among them.
Such a theory avoids miscellany, the paradigm of which is a mere
list, which says nothing about how, if at all, its various items are
ordered or organized.
Sixth, the theory is coherent, not only internally but also exter-
nally, fitting well with a wide range of understanding-­providing
theories of other domains. A theory of one subject matter that
massively conflicts with a coherent cluster of accurate, reason-­
based, robust, illuminating, and orderly theories of other domains
does not further comprehension but muddles it (yet another char-
acteristic of misunderstanding).
Although all six properties contribute to theoretical under-
standing, they do so in different ways. The latter two, unlike the
former four, only conditionally make such contributions. The
orderliness and coherence of a theory contribute to its ability to
supply understanding only if the theory possesses the other four
features to at least some extent. In this way, these first four are fun-
damental to understanding in a way that the final pair are not.
When inquirers fully grasp theories with these six properties,
the targets of their theories make sense to them. This is theoretical
understanding, which on our construal is distinctive. It isn’t re­du­
cible to true or justified belief (or to their conjunction). Nor is it
30 The Nature of Inquiry

equivalent to ordinary knowledge. For, as explained above, the

function of such understanding is, inter alia, to illuminate, in a
robust, orderly, coherent fashion, the portion of reality under
investigation, and not simply to state a series of known truths or
justified beliefs about it. Theoretical understanding is not epi­
stem­ic perfection. Still, it is a type of epistemic excellence. And it
is a prime candidate for an ultimate proper goal of inquiry.20
To appreciate this last claim, notice that it handles the points
that challenged other aspirants. Theoretical understanding could
not be unwarranted or narrowly circumscribed, and is neither
dissonance-­inducing nor compatible with obliviousness. It also
excludes the sorts of confusion and muddle illustrated by our
examples involving ratio and social construct. That it ticks
these boxes is unsurprising. After all, were inquirers to fully
grasp a theory that is accurate, reason-­based, robust, illuminating,
orderly, and coherent, their grip on the inquiry’s subject matter
would be both comprehensive and systematic. They would pos-
sess accurate replies to a wide range of central questions about
that subject matter, including explanatory ones. Their answers
would fit together, and enjoy rational support. Moreover, the
inquirers would grasp those answers without conceptual confu-
sion. In such a case, their curiosity, wonder, or puzzlement would
be appropriately relieved.

20
Recent epistemology has witnessed a surge of arguments for the value of under-
standing over all other epistemic achievements: see, among others, Elgin (1996, 122ff.),
Zagzebski (2001), Kvanvig (2003, Ch. 8), and Pritchard (2010, Ch. 4 and 2016, §§2–4);
cp. Dellsén (2016). We will not rely on these arguments here, though they may offer
independent support for our assessment. See Bengson (2017) for a fuller treatment of
understanding.
 Theoretical Understanding 31

This line of reasoning may be buttressed by observing that a

wide range of more demanding epistemic or alethic achievements,
such as absolute certainty, are unnecessary to obtain such relief.
Such achievements might be desirable in various respects. But
even were inquirers to fall short of them, that wouldn’t by itself
impugn the success of their inquiries, so long as they terminated
in full grasp of theories with the six properties we’ve identified.
Our arguments in the previous section also cast doubt on the
claim that there is a less demanding epistemic or alethic achieve-
ment that is an ultimate proper goal. Though we certainly haven’t
offered a decisive refutation of this claim, we have made the case
that ordinary knowledge fails to resolve inquiry. (We’ve also given
reason to believe that less demanding states, such as true belief,
justified belief, and justified true belief, fail as well.) So any ultimate
proper goal of inquiry distinct from understanding would have to
be stronger than knowledge, but at least as strong as understand-
ing. We’re not sure what that might be—especially given our con-
tention above that anything stronger than understanding is
unnecessary. In any event, there is an important sense in which
this doesn’t matter. For we’ve provided reason to think that under-
standing is the minimal epistemic achievement needed to resolve
inquiry. And that’s all our arguments in this book require.
(In order to streamline our discussion, we’ll often write as if the-
oretical understanding is inquiry’s ultimate proper goal. This
should not be read as implying that there aren’t any other ultimate
proper goals, but rather as shorthand for the claim that under-
standing suffices to resolve inquiry, and no weaker epistemic or
alethic achievement can do so.)
The special role we are attributing to understanding can be fur-
ther motivated through a contrast with the popular idea—recently
32 The Nature of Inquiry

defended by David Chalmers—that inquirers ought to aim for

collective convergence on the truth. Chalmers contends that other
pos­sible goals, including understanding, involve “something of a
lowering of our sights.”21 While we endorse the emphasis Chalmers
places on collectivity, we reject a narrow focus on truth and agree-
ment. In addition to concerns raised earlier (in §2) about the rela-
tion between true belief and inquiry’s ultimate proper goal, we’d
like to register two further worries about Chalmers’ proposal.
Both rely on drawing attention to the six understanding-­enabling
properties we’ve identified.
First, collective convergence on the truth appears to be insuffi-
cient to resolve inquiry. For such convergence would not ensure pos-
session, let alone full grasp, of a theory that possesses all six of the
features described above. But in the absence of this, we could hardly
deem inquiry complete. There would remain more work to do.
Second, convergence is unnecessary, and this is so whether it
consists in full or partial agreement. Suppose we discovered that
human psychology (with its familiar penchant for generating dis­
agree­ment) rendered convergence unattainable. Still, were in­quir-
ers fortunate enough to develop a theory that realized all six
features associated with theoretical understanding, fully grasping
that theory would suffice to resolve inquiry, even if agreement
remained elusive.
We conclude that Chalmers’ charge gets things backwards.
Convergence on the truth closes inquiry only when it is ac­com­pan­
ied by understanding. In this sense, it is aiming for mere agreement
on the truth that would be “something of a lowering of our sights.”

21
Chalmers (2014, 14). By ‘the truth,’ Chalmers is referring to the proposition that
correctly answers the question under investigation.
 Theoretical Understanding 33

That said, as anticipated above, we support Chalmers’ suggestion

that inquiry’s aspiration be couched in collective terms. Specifically,
we affirm that an ultimate proper goal of inquiry must be both
capable of being pursued collectively and available to collectives.
Theoretical understanding meets these conditions. It can be
realized through collective efforts at data-­gathering, the­or­iz­ing,
or both. And the understanding that results from those efforts is
in principle widely available.
We’ve been focusing on the importance of understanding to the-
oretical inquiry. It may be that science, philosophy, and neighbor-
ing disciplines incorporate other, non-­ epistemic projects as
well—e.g., those that are aesthetic, felicific, salvific, pragmatic, or
political. No doubt philosophers have at times appraised compet-
ing views by these (and still other) non-­epistemic standards. Our
intention is not to question the legitimacy of such appraisals, or to
promote a picture of inquiry on which its sole interest derives from
the ambition of attaining theoretical understanding with respect to
its central questions. Rather, our contention is simply that such
understanding is theoretical inquiry’s ultimate proper goal.
This contention should be congenial to those with disparate
metaphilosophical commitments. Granted, many pragmatists,
intellectual anarchists, and anti-­theorists appear to decry the sorts
of ambitions we have set for inquiry.22 But their misgivings are

22
All three positions have long been thought to have severe metaphilosophical
implications. For two influential works by anti-­theorists, see Baier (1985, Ch. 12) and
Williams (1985). Dewey (1910 and 1938) is a well-­known pragmatist, and Feyerabend
(1975) famously avows anarchism. Urmson (1967, 232) also gives voice to the anarchist
view in his summary of J. L. Austin’s work: “No one knows what a satisfactory solution
to such problems as those of free will, truth, and human personality would look like,
and it would be baseless dogmatism to lay down in advance any principles for the
proper method of solving them.”
34 The Nature of Inquiry

typically directed at ideas to which we are not committed or that

we explicitly reject, such as an exaggerated sense of the im­port­
ance of simplicity or an inflated notion of objectivity as requiring
an “absolute conception of the world.” Moreover, our approach
coheres with the stated motivations of many pragmatists, an­arch­
ists, and anti-­theorists. These motivations include the historical
and cultural situatedness of reflective thought, and the wide var­
iety of considerations relevant to such reflection. Our contention
is also compatible with the position that, in some cases, the­or­et­ic­al
understanding with respect to certain questions is not achievable,
or at least is not readily available. We agree, for instance, with
David Wiggins when he writes that it is often “a matter of prolonged
and difficult inquiry gradually to improve currently accepted
standards or conceptions” in philosophy.23
If our contention is correct, two conclusions follow. First, a
method is sound only if it possesses the resources to guide the
construction of a theory that positions inquirers to acquire under-
standing, and (concomitantly) to facilitate evaluations of candi-
date theories in this light. Second, because methods take data as
their input, we must get clear about the nature of data and the
means by which they are to be collected. We devote the next two
chapters to presenting views about data in philosophy that are
consonant with their role in developing a philosophical method
poised to yield successful theories. We’ll then devote the remain-
der of the book to assessing extant methods, presenting the
method that we prefer, and revealing its many benefits.

23
Wiggins (2001, 82).