Berkeley’s Theory of Language∗

Kenneth L. Pearce
Trinity College Dublin
January 7, 2019

Abstract
According to the ideational theory of meaning held by most 17th and
18th century European philosophers, meaningful words signify ideas in the
mind of the speaker. This view was famously rejected by George Berkeley.
Some interpreters hold that, in rejecting this view Berkeley meant merely
to carve out some exceptions to the general principle that meaningful
words stand for ideas. Others attribute to Berkeley the view that, rather
than signifying ideas, words signify speaker intentions. Finally, some in-
terpreters attribute to Berkeley a use theory of language similar to the
later Wittgenstein. This article provides an overview of this interpretive
debate and a defence of the use theory interpretation.

Keywords: George Berkeley; philosophy of language; signification; ideational

theory of meaning; speaker intentions; Ludwig Wittgenstein; use theory
of meaning

In the Introduction to the Treatise concerning the Principles of Human

Knowledge, Berkeley attacks the “received opinion that language has no other
end but the communicating our ideas, and that every significant name stands
for an idea” (PHK, Intro 19:37). How far does Berkeley go in rejecting this
‘received opinion’ ? Does he offer a general theory of language to replace it? If
so, what is the nature of this theory?
The ‘received opinion’ Berkeley opposes is often known as the ‘ideational
theory of meaning’ because it holds that a word gets to be meaningful by signi-
fying an idea in the mind of the speaker. The aim of language is to communicate
this idea to the hearer. §1 outlines this view in more detail.
Many interpreters have thought that, in rejecting the claim that language
has no other end than communicating ideas and that every significant name
stands for an idea, Berkeley meant to be carving out one or more exceptions to
the ideational theory of meaning. I call this the ‘modified ideational theory’. In
§2 I argue that the modified ideational theory cannot be Berkeley’s view: after
∗ Invited contribution to The Oxford Handbook of Berkeley, ed. Samuel C. Rickless (Oxford

University Press).

1
Berkeley’s Theory of Language 2

all the ‘exceptions’ to the ideational theory Berkeley carves out, there is nothing
left of the theory at all.
A second interpretation of Berkeley’s philosophy of language holds that lan-
guage serves to communicate the intentions of the speaker. In §3 I argue that,
although this account solves many of the problems associated with the modified
ideational theory, it too must ultimately be rejected.
A third interpretation attributes to Berkeley a use theory of meaning, similar
to the later Wittgenstein. In §4, I defend this interpretation: Berkeley’s view
is that a word gets to be meaningful when it is used according to conventional
rules within a community to accomplish some practical purpose.

1 The Ideational Theory of Meaning

According to Aristotle, “spoken sounds are symbols of affections of the soul”
(De Int 1 16a4). For most early modern philosophers, beginning from Descartes,
these ‘affections of the soul’ are, primarily, ideas. Thus Descartes writes, “when-
ever I express something in words, and understand what I am saying, this very
fact makes it certain that there is within me an idea of what is signified by the
words in question” (CSM, 2:113).
This approach to the philosophy of language was worked out in detail in the
Port-Royal Grammar (Arnauld and Lancelot [1660] 1975) and Logic (Arnauld
and Nicole [1662] 1996). Although the Port-Royalists often say, as Descartes
did, that all meaningful words signify ideas (e.g., Arnauld and Nicole [1662]
1996, 66), their considered view is that meaningful words “indicate what takes
place in the mind” of the speaker (Arnauld and Nicole [1662] 1996, 74). This
includes signifying both “the objects of our thoughts” (i.e., ideas) and “the
manner or mode of our thoughts” (Arnauld and Lancelot [1660] 1975, 67–68).
The most important ‘manner or mode’ of thought is judgment, which is signified
by the verb (Arnauld and Lancelot [1660] 1975, 122; Arnauld and Nicole [1662]
1996, 79). Judging occurs when, “After conceiving things by our ideas, we
compare these ideas and, finding that some belong together and others do not,
we unite or separate them” (Arnauld and Nicole [1662] 1996, 82). The act of
judging forms a proposition, that is, a truth-evaluable mental state with subject-
predicate structure. When a complete sentence is used meaningfully, it signifies
this complex mental state in the mind of the speaker—that is, it makes this
state known to the hearer.1
Essentially the same approach is adopted by Locke. According to Locke,
words are “articulate Sounds . . . use[d] . . . as Signs of internal Conceptions”
(E, 3.1.2:402). Like Port-Royal, Locke usually says that all meaningful words
stand for ideas, and argues at some length that “Words in their primary or
immediate Signification, stand for nothing, but the Ideas in the Mind of him
that uses them” (E, 3.2.2:405).2 However, also like Port-Royal, Locke carves out
1. For a more detailed overview of the Port-Royal theory of language, see Pearce 2015.
2. On the interpretation of this claim, and Locke’s argument for it, see Ashworth 1981,
(1984) 1998; Guyer 1994, §2; Losonsky 1994, 2007; Pritchard 2013.
Berkeley’s Theory of Language 3

an important exception for the class of words he calls ‘particles’. These words,
Locke says, “signify the connexion that the Mind gives to Ideas, or Propositions,
one with another.” Included in this class are “Is, and Is not, [which] are the
general marks of the Mind, affirming or denying” (E, 3.7.1:471). Locke also
follows Port-Royal in taking affirmation and denial to involve “The joining or
separating of signs” to construct a truth-evaluable proposition (E, 4.5.2:574).
Locke distinguishes sentences, which he calls ‘verbal propositions’, from the
corresponding mental states, which he calls ‘mental propositions’ (E, 4.5.3:574).
The mental proposition, like the verbal proposition, exhibits subject-predicate
structure.
The ideational theory, endorsed by Port-Royal and by Locke, holds that the
human mind possesses a variety of ideas and the capacity to perform acts of
judgment whereby these ideas are assembled into truth-evaluable states with
subject-predicate structure.3 Words are conventional signs meant to signify
these ideas and judgments, that is, to let the hearer know what the speaker is
thinking. This process, however, is error-prone, and many of our mistakes in rea-
soning actually stem from the various ‘imperfections’ and ‘abuses’ of language.
These failings all come down to the problem of developing a stable association
between a word and some one single idea (Arnauld and Nicole [1662] 1996,
58–66; E, 3.9.4–5:476–477, 3.10.2–3:490–491, 3.10.5:492–493, 3.10.28:505–506).
Uncertainty, inconstancy, or imprecision regarding the idea signified by a word
leads to failures of communication and, as Berkeley observed, communicating
ideas is, according to this ‘received opinion,’ “the chief and only end of language”
(PHK, Intro 20:37).

2 The Modified Ideational Theory

In the manuscript and published versions of the Introduction to the Principles
and in Alciphron, Berkeley quite explicitly argues against the ideational theory
of meaning. His presentation of the view he is attacking looks very much like
the theory of Port-Royal and Locke outlined in the previous section. Thus in
the published Introduction Berkeley’s opponents are said to hold “that language
has no other end but the communicating our ideas, and that every significant
name stands for an idea” (PHK, Intro 19:37), and in Alciphron the freethinking
title character (representing Berkeley’s opponents) says:
Words are signs: they do or should stand for ideas; which so far as
they suggest they are significant. But words that suggest no ideas are
insignificant. . . He who really thinks has a train of ideas succeeding
each other and connected in his mind: and when he expresses himself
by discourse, each word suggests a distinct idea to the hearer or
reader; who by that means has the same train of ideas in his, which
was in the mind of the speaker or writer. As far as this effect is
3. For detailed treatments of judgments and propositions in Port-Royal and Locke, see
Mattern (1978) 1998; Buroker 1996; Ott 2002; Schaar 2008; Rickless 2008; Marušić 2014.
Berkeley’s Theory of Language 4

produced, so far the discourse is intelligible, has sense and meaning.

(Alc, 7,2:287–288)4
It is undisputed that Berkeley means to reject this view. The question at
issue is how precisely he means to reject it. According to what has perhaps been
the most popular interpretation, Berkeley does not reject the ideational theory
outright, but instead carves out a series of exceptions to it. As John Russell
Roberts has memorably put it, the proponents of this interpretation hold that
“Berkeley is cutting off only a sickly limb of Locke’s view rather than digging
up the whole by its roots” (Roberts 2007, 43). I call this type of interpretation
a ‘modified ideational theory’.
Any interpretation of Berkeley must allow that ideas sometimes play some
role in securing the meaningfulness of words. Thus a modified ideational theory
is not just any interpretation of Berkeley’s philosophy of language that some-
how involves ideas. Rather, a modified ideational theory is an interpretation
which attributes to Berkeley the view that there is some significant fragment of
language to which the original ideational theory applies. This view is endorsed
explicitly by several interpreters, including Bennett (1971, §10), Hacking (1974,
15–16, 36–39), and Bordner (2017, 263), and appears to be assumed implicitly
by many others. In this section, I argue that no modified ideational theory is a
viable interpretation of Berkeley.
There are at least two areas of language in which Berkeley very explicitly
allows for the meaningful use of words without ideas. First, Berkeley holds that
since
an idea [is] altogether inactive. . . An agent. . . an active mind, or
spirit, cannot be an idea or like an idea. Whence it should seem
to follow that those words, which denote an active principle, soul,
or spirit, do not in a strict and proper sense, stand for ideas: and
yet they are not insignificant neither. (Alc, 7,5:292)
Second, Berkeley allows that in addition to the communication of ideas,
There are other ends [of language], as the raising of some passion,
the exciting to, or deterring from an action, the putting the mind in
some particular disposition; to which the former [i.e., communication
of ideas] is in many cases barely subservient, and sometimes entirely
omitted, when these can be obtained without it. (PHK, Intro 20:37,
emphasis added)
Berkeley’s view on this latter category of language has often been called his the-
ory of ‘emotive meaning’ (e.g., Belfrage 1986a; 1987a, §3; Berman 1994, 12–17,
4. In this passage, Berkeley interprets the ideational theory as holding that an utterance
is understood just when it produces in the mind of the hearer a copy of the speaker’s mental
state. Jennifer Marušić (2014, 273–277) has suggested, with considerable plausibility, that in
at least some cases understanding an utterance may instead require the hearer to have an
idea of the speaker’s mental state. Berkeley’s critique of the ideational theory does not rely
on any particular position on this issue.
Berkeley’s Theory of Language 5

144–159). However, as we shall see, there is considerable dispute about how

similar Berkeley’s theory here is to 20th century emotivism. Kenneth Williford
(2003) has therefore proposed the more neutral label ‘operative language’, based
on a remark in Alc, 7,14:307. According to the modified ideational interpreta-
tion, Berkeley holds that the ideational theory gives the correct account of words
like ‘apple’ (PHK, 1:41) but a different account is needed for operative words
like ‘good’ (PHK, Intro 20:37) and spirit words like ‘God’ (DHP, 231–232).
The first problem facing the modified ideational theory is to make sense of
the category of operative language. Berkeley clearly includes ethical language
and the language of religious ‘mysteries’ such as the Trinity in this category, as
well as words like ‘reward’ and ‘danger’ which serve primarily to shape feelings
and behavior (MI, 36–43:105–113). However, this is not a matter of a certain
special class of words: Berkeley explicitly says that even the sentence ‘Aristotle
hath said it’ may function ‘operatively’ if “all . . . [the speaker] means by it is to
dispose [the hearer] to embrace his opinion with the deference and submission
which custom hath annexed to that name” (PHK, Intro 20:38). The modified
ideational theory requires some kind of account of the scope of the operative
language exception to the ideational criterion of meaningfulness, but it must
also recognize that, even in operative language, the communication of ideas is
only “sometimes entirely omitted” (PHK, Intro 20:37, emphasis added). The
word ‘reward’ can simultaneously function operatively and signify a particular
ice cream cone idea (see Pearce 2017c, 40). Thus, contrary to Bennett (1971,
53–54) and Berman (1994, 148), it does not appear that Berkeley introduces a
demarcation between ‘cognitive’ and ‘non-cognitive’ discourse.5
This demarcation problem is not the only difficulty the modified ideational
theory faces in making sense of operative language. The language of religious
mysteries is an undisputed example of language that Berkeley regards as ‘op-
erative’, yet Berkeley argues at length that such religious language gets to be
meaningful in precisely the same way as the language of physics and algebra
(Alc, 7,5–8:291–299). Thus, if operative language is non-cognitive language,
then not only is Berkeley, bizarrely, a non-cognitivist about Aristotle ascrip-
tions, he is also a non-cognitivist about math and physics (Williford and Jakapi
2009, 106). Additionally, Berkeley holds that operative utterances can be true
or false and stand in entailment relations (Alc, 7,7:295, 7,9:300–301; see Jakapi
2002, 2007). ‘Non-cognitivism’ about a domain of discourse is often defined as
the thesis that the utterances in that domain are neither true nor false. On
standard non-cognitivist theories, such as the emotivism of A. J. Ayer ([1946]
1952, 107–109), which Berman (1994, 155) says is the same as Berkeley’s, non-
cognitive words stand for ‘mere pseudo-concepts’ and lack the sort of semantic
content needed for sentences containing them to be true or false or stand in
entailment relations. This conflicts with what Berkeley says about religious
mysteries (see Pearce 2017c, 61).
It might be thought that this last difficulty is not essential to the modi-
fied ideational theory and is merely the product of an anachronistic attempt
5. For further argument in support of this claim, see Roberts 2007, 58–65.
Berkeley’s Theory of Language 6

to fit Berkeley into early 20th century debates. On the contrary, this is not
merely a product of Berman’s assimilation of Berkeley to Ayer, but rather a
general problem that must be confronted by any modified ideational interpre-
tation. Port-Royal and Locke made essential use of their theory of the mental
proposition in their accounts of truth and inference. However, a proposition,
on the ideational theory of meaning, is a joining or separating of two ideas. If
the subject term or the predicate term in a sentence does not stand for an idea,
then the ideational theory’s account of truth and inference cannot be employed
and some other account must be sought.
In fact, the difficulty is even more serious than this, and brings us to the
second main problem for the modified ideational interpretation: Berkeley’s re-
jection of abstract ideas prevents him from accepting the ideational theory’s ac-
count of the meaning of even simple subject-predicate sentences like ‘the apple is
red’ or (to use Berkeley’s own example) ‘Melampus is an animal’. However, this
is just the portion of language with respect to which proponents of the modified
ideational interpretation claim that Berkeley has preserved Locke’s theory. If
Berkeley cannot endorse what the ideational theory says about these sentences,
then the exceptions have completely swallowed the rule and there is nothing left
of the ideational theory at all.
In his discussion of the sentence ‘Melampus is an animal’ in the Manuscript
Introduction, Berkeley correctly reports that, according to the ideational theory
espoused by Port-Royal and Locke, the word ‘Melampus’ in this sentence stands
for “the Idea of some one particular. . . Dog” while “the Name Animal stand[s]
for an Abstract, Generical Idea” of animal (MI, 34:101–103). The word ‘is’
stands for the mind’s act of judging that these two ideas agree.
Berkeley clearly cannot endorse this account since he argues that ‘Abstract,
Generical Ideas’ do not exist. However, it might be thought that only a very
minor modification is required since, in the published Introduction, Berkeley
accepts the existence of general ideas, provided these ideas are not understood
to be abstract.6 Thus Berkeley might still hold that the word ‘animal’ in the
sentence signifies a general idea.
In fact, Berkeley cannot endorse this moderate position. When Berkeley
admits the existence of general ideas in the published Introduction to the Prin-
ciples, he writes:

as the [general idea of a line] owes its generality, not to its being
the sign of an abstract or general line, but of all particular right
lines that may possibly exist, so the [word ‘line’] must be thought
to derive its generality from the same cause, namely, the various
particular lines which it indifferently denotes. (PHK, Intro 12:32)

Since general words and general ideas get to be general in just the same way
it cannot be the case that general words get to be general by standing for
general ideas, as Locke supposed (E, 3.3.6:410–411). If this were the case, then
6. In the manuscript, Berkeley had rejected general ideas entirely (MI, 20:83, 27:93; see
Belfrage 1986b).
Berkeley’s Theory of Language 7

Berkeley would have to say that general ideas get to be general by standing for
other general ideas, generating a regress.
It might be thought that Berkeley’s view here is that, each time we meaning-
fully use or understand the word ‘animal’, we have some one of the ideas in its
extension in mind (but not always the same one).7 However, Berkeley explicitly
rejects this view in the 1732 edition of Alciphron (7,7*:335) and he presents an
argument against it in the ‘Melampus’ passage of the Manuscript Introduction.
His argument is as follows:
in the Proposition we have instanced [i.e., the sentence ‘Melampus is
an animal’] in it is plain the Word Animal is not suppos’d to stand
for the Idea of any one particular animal. for if it be made stand
for another different from that is marked by the Name Melampus,
the Proposition is false and includes a Contradiction. And if it be
made signify the very same Individual that Melampus doth, it is a
tautology (MI, 35:105, crossed out and bracketed text omitted).
‘Melampus is Melampus’ is a tautology, and ‘Melampus is Fido’ is a false-
hood, but ‘Melampus is an animal’ is neither tautologous nor false. Without
abstract ideas, the ideational theory of meaning cannot explain this fact (see
Pearce 2017b, 566–567; 2017c, 143–144). Berkeley therefore concludes, con-
trary to the ideational theory, that “in that Proposition [the word ‘animal’ does
not] stand for any Idea at all. All that I intend to signify thereby being only
this. That the particular thing I call Melampus has a right to be called by the
Name Animal” (MI, 34:103, bracketed text omitted). In this context, there is no
idea corresponding to the word ‘animal’. Consequently, there can be no mental
proposition corresponding to the sentence ‘Melampus is an animal’.
In the notebooks, Berkeley expressed puzzlement about the absence of men-
tal propositions in certain cases of apparently meaningful speech (PC, 728–
731a:96, 738:97, 793:104, 809:106). By the time he wrote the ‘Melampus’ dis-
cussion in the Manuscript Introduction, he had rejected the mental proposition
entirely.
It may be objected that my argument has relied heavily on manuscript ma-
terial and on a comment in Alciphron that Berkeley deleted from later editions.
I concede that this provides a good reason for us to wonder whether Berkeley
might have rejected, or at least been uncertain about, the views in question.
However, alternative interpretations still need to explain why Berkeley wrote
these words in the first place, even if he later rejected them. The discussion of
propositions in the Manuscript Introduction shows that from the beginning of his
career Berkeley was aware of a compelling argument for the claim that someone
who rejects abstract ideas cannot accept mental propositions, and nowhere does
Berkeley do anything to undermine that argument. Without abstract ideas, the
theory of mental propositions is untenable, and without mental propositions the
ideational theory of meaning, together with the associated theories of truth and
inference, collapses.
7. Melissa Frankel (2009, 383n5) suggests that PHK, Intro can be read this way, though
she recognizes that Berkeley rejects this view in Alciphron.
Berkeley’s Theory of Language 8

As soon as he rejected abstract ideas (c. 1708),8 it became impossible for

Berkeley to accept any significant fragment of the ideational theory of meaning,
even restricted to some limited domain.
Berkeley thought Port-Royal and Locke were radically mistaken about the
relationship between words and ideas, and he argues quite explicitly that words
may sometimes be meaningful in the absence of ideas. Nevertheless, Berkeley
continues to speak of associations between words and ideas in at least some
contexts (e.g., NTV, 32:181–182, 51:190, 152:232–233; PHK, Intro 21–25:38–
40; PHK, 16:47, 43:58). Berkeley therefore stands in need of an alternative
theory of how words get to be meaningful, and this theory should explain why
and how meaningfulness sometimes involves ideas and sometimes does not. In
the remainder of this chapter, we consider two alternative interpretations of
Berkeley’s theory.

3 The Speaker Intentions Theory

In the Manuscript Introduction, Berkeley writes:
It is plain therefore that a Man may understand what is said to him
without having a clear and determinate Idea annexed to and marked
by every particular [ Word ]9 in the Discourse he hears. Nay, he may
perfectly understand it. For what is it I pray to understand perfectly,
but to understand all that is meant by the Person that Speaks? (MI,
43:113)
Berkeley here speaks of what is meant by the speaker. Elsewhere, in similar
contexts, he speaks of what is intended by the speaker. For instance, in the
manuscript version of the ‘Aristotle’ passage, he writes, “wn a Schoolman tells
you that Aristotle hath said it, think you that he intends to excite in your
Imagination the Idea of that particular Man? All He means by it is only to
dispose you to receive his Opinion” (MI, 38:109). In these and other passages of
the Manuscript Introduction, what the speaker means appears to be identified
with the effect the speaker intends to produce, and understanding a speaker’s
utterance appears to be a matter of grasping the speaker’s intention.
According to Kenneth Williford (2003), this forms the core of Berkeley’s
theory of language.10 Williford argues that, in the Manuscript Introduction,
Berkeley
suggests that the intended end for which an utterance is put to
use, i.e. the speaker’s intention in uttering, partially determines the
meaning of an expression. Second and correspondingly, he suggests
that an integral part of understanding an utterance is understanding
what the speaker intends to accomplish with it (Williford 2003, 277).
8. On the dating of MI, see Belfrage 1987b, 20–23. On the question of when precisely
Berkeley rejected abstract ideas, see Belfrage 1986b.
9. Brackets in original.
10. A similar interpretation is hinted at by Winkler 2009, 495–496.
Berkeley’s Theory of Language 9

Williford here indicates that speaker intentions are only part of the story. Later
in the paper he says that “If the speaker intends to communicate ideas, then
there is a further step involved [in understanding the utterance]: [the hearer]
must then also grasp the ideas answering to the terms of the speaker’s sentence”
(Williford 2003, 287). The point of Berkeley’s discussion of operative language
is to argue that this is not always part of the speaker’s intention.
The speaker intentions theory, like the modified ideational theory, sees Berke-
ley as agreeing with the view of Port-Royal and Locke that words “indicate what
takes place in the mind” of the speaker (Arnauld and Nicole [1662] 1996, 74).
According to this interpretation, Berkeley’s innovation lies in considering the
relevant items in the mind of the speaker to be intentions rather than ideas.
This approach has a number of advantages over the modified ideational
theory. First, it preserves Berkeley’s claim that ‘operative’ religious utterances
get to be meaningful in the just the same way other utterances do: they serve
to communicate the intentions of the speaker. Second, this view accounts nicely
for Berkeley’s claim about the ‘Melampus’ sentence. Berkeley writes, “All that
I intend to signify [by the sentence ‘Melampus is an animal’] being only this.
That the particular thing I call Melampus has a right to be called by the Name
Animal” (MI, 34:103, bracketed text omitted). The effect the speaker intends
to produce here is that the hearer should come to regard the application of the
label ‘animal’ to Melampus as appropriate.
Nevertheless, the speaker intentions theory does not ultimately succeed in
making sense of Berkeley’s remarks about language. First, the speaker inten-
tions theory does not provide a plausible account of operative language. In the
published Introduction, Berkeley writes:
I entreat the reader to reflect with himself, and see if it does not often
happen either in hearing or reading a discourse, that the passions of
fear, love, hatred, admiration, disdain, and the like, arise immedi-
ately in his mind upon the perception of certain words, without any
ideas coming between. (PHK, Intro 20:37)
Shortly after the manuscript version of this passage, Berkeley writes, “Upon
hearing the Words Lie Rascal, Indignation, Revenge and the suddain motions
of Anger do instantly [ ensue ]11 in the minds of Some Men without ever at-
tending to the Definition of those Names” (MI, 42:113, some crossed out and
bracketed text omitted). It seems unlikely that Berkeley thinks that the word
‘danger’ produces fear by making the hearer conceive that the speaker intends
her to be afraid, or that the word ‘rascal’ produces anger by making the hearer
conceive that the speaker intends her to be angry. Rather, these passions arise
‘immediately’ or ‘instantly’ and are not mediated by notions about the speaker’s
intentions any more than they are mediated by ideas.
Williford argues that his interpretation is supported by the final stratum of
the manuscript’s discussion of moral language. This is a brief but dense text
and Berkeley marked it up extensively before crossing out the entire section and
11. Brackets in original.
Berkeley’s Theory of Language 10

omitting any discussion of moral language in the published version. It is worth

pausing to examine it in some detail.
In the first stratum of the manuscript, the relevant portion reads:
I ask any Man whether when he tells another that Such an Action is
Honourable and vertuous, he has at that instant the Abstract Ideas
of Honour and Vertue in his [thoug]12 view, and whether in reality
his intention be to raise those abstract ideas together with their
agreement to the particular Idea of that Action in the Understanding
of him he speaks to. Or rather, whether this be not his full Purpose
namely that those Words should excite in the Mind of the Hearer an
esteem of that particular Action and stirr him up to the performance
of it? (MI, 41:111)
As Williford acknowledges, this is the closest thing to an explicit endorsement
of emotivism in Berkeley’s corpus (Williford 2003, 296). However, Berkeley was
clearly unsatisfied with this formulation. Before rejecting the passage entirely,
he modifies it to read:

I ask any Man whether every time he tells another that Such an
Action is Honourable & vertuous with an intention to excite him
to the performance of it he has at that instant Idea of Honour &
Vertue in his view, and whether in reality his intention be to raise
Idea together with their agreement to the idea of that particular
Action in the Understanding of him he speaks to.
In this later stratum of the manuscript, Berkeley is careful to speak, not of the
sentence in general, but of the speaker’s intentions on a particular occasion.
Williford argues, with considerable plausibility, that Berkeley’s revisions are in-
tended to avoid the implication that moral words never stand for ideas (Williford
2003, 296–297). Again, however, it is implausible to suppose that grasping the
speaker’s intention “to excite him to the performance of” the action in question
will motivate the hearer. Rather, as the concluding sentence common to the
earlier and later strata indicates, this use of moral language “stirr[s the hearer]
up to the performance of” the action by “excit[ing] in the Mind of the Hearer
an esteem of that particular action.” This, again, is an immediate effect of the
words which need not be mediated by either ideas or notions.
This difficulty is closely connected with another problem that Williford
briefly recognizes (Williford 2003, 288n31), which we might call ‘the problem
of hidden agendas’. Not only is it often the case that the hearer’s grasping the
speaker’s intentions would not assist in realizing those intentions, sometimes it
would actively prevent the speaker from fulfilling those intentions. This will
happen in any case of deception or manipulation.
There is also a difficulty about the speaker intentions theory’s relationship
to the theory of notions. Speaker intentions are active, and hence there can
12. Brackets in original.
Berkeley’s Theory of Language 11

be no ideas, but only notions, of a speaker’s intentions. In a more recent pa-

per Williford, in collaboration with Roomet Jakapi, suggests that in Alciphron
words like ‘force’ and ‘grace’, which (unlike the examples found in the early
works) never signify ideas, signify systems of relations (Williford and Jakapi
2009). Relations, according to Berkeley’s mature view, are also known by no-
tions and not ideas (PHK, 89:80, 142:106, [1734 ed.]). However, Berkeley says
very little about what it is to have a notion of, e.g., spirit, other than that it is
to understand the word ‘spirit’ (PHK, 27:53 [1734 ed.]). Some interpreters have
argued on this basis that, for Berkeley, to have a notion just is to understand a
word (Tipton 1974, 269–271; Woozley 1976; Pearce 2017c, 125–126). Williford
and Jakapi’s interpretation requires that we be able to think about the inten-
tions of others and about systems of relations in a way that is prior to language,
but there is reason to doubt that Berkeley’s account of notions allows this.13
In fact, in the 1752 edition of Alciphron Berkeley explicitly asserts that “rela-
tions, habitudes or proportions. . . cannot be by us understood but by help of
signs” (Alc, 7,14:307). Our grasp of relations is dependent on our ability to use
words or other signs, and not prior to it, as Williford and Jakapi’s interpretation
requires.14
The final and most serious difficulty for the speaker intentions theory is that
Berkeley does not merely argue that words can be used without ideas to influ-
ence feelings and actions, he argues that “general names are often used in the
propriety of language without the speaker’s designing them for marks of ideas”
(PHK, Intro 20:38, emphasis added). ‘The propriety of language’ refers to the
use of words in accord with the rules of a language, as opposed to idiosyn-
cratic uses in unusual circumstances or by unusual speakers (Pearce 2017a, 237,
248n16; compare Williford 2003, 280n18). What this shows is that Berkeley
does recognize a distinction between what a speaker means by a particular ut-
terance on a particular occasion and the use of those words which is authorized
by the rules of language. Regarding the latter, Berkeley affirms that “propriety
[is] regulated by custom” (PHK, 52:63; also see DHP, 216, 250). What Berkeley
is trying to show in the manuscript and published introductions is not simply
that speakers sometimes use words without intending to mark ideas, but that
this use is authorized by our linguistic conventions.
In Alciphron the status of the rules of propriety as our (plural) conventions
is much clearer. In that text, Euphranor sets out the central problem of the
philosophy of language as follows: “Be the use of words or names what it will,
I can never think it is to do things impossible [i.e., to suggest abstract ideas].
Let us then inquire what it is? and see if we can make sense of our daily
practice” (Alc, 7,5:291, emphasis added). In the discussion that follows, the
use of language is compared to a poker game, and this is followed by a series
13. This problem is briefly recognized by Williford (2003, 287n30). Daniel Flage (1985)
defends an interpretation of notions as pre-linguistic mental acts. This interpretation, if
successful, would solve Williford and Jakapi’s circularity problem.
14. Williford and Jakapi (2009, 113) explicitly recognize that relations cannot be grasped
apart from signs, but they give no account of how, within their theory, it is possible to grasp
relations by means of signs.
Berkeley’s Theory of Language 12

of reflections on not only the individual but also the social impact of various
religious and scientific forms of words.
I conclude, therefore, that Williford has made a strong case for the im-
portance of speaker intentions to Berkeley’s understanding of speaker meaning.
However, there are a number of reasons why this cannot be the whole story even
regarding speaker meaning. Further, the essential publicity of language is rec-
ognized in Berkeley’s early philosophy and emphasized in his later philosophy.15
The speaker intentions theory is not, however, a theory of ‘the propriety of lan-
guage’ and thus is not an adequate answer to the central question of philosophy
of language as Berkeley himself framed it.

4 The Use Theory

4.1 Meaning as Use in Alciphron
The treatment of language in Alciphron VII, as has been observed, begins by
characterizing language as a public social practice in which we are all immersed,
and asks how words are used within this practice. “Words, it is agreed, are
signs” and therefore, Euphranor suggests, in understanding their use “it may
not . . . be amiss to examine the use of other signs, in order to know that of
words” (Alc, 7,5:291). The first example is that of poker chips. The chips
(‘counters’), the characters agree, get to stand for money because they are
purchased at the beginning and cashed out at the end. In between, they are
simply used according to the rules and it doesn’t matter whether the players
ever think about the money or not.
This leads to a discussion of the rules according to which various problematic
words—most notably ‘grace’ as a technical term of theology and ‘force’ as a
technical term of physics—are used in their respective discourses, and the ends
at which these rule-governed uses aim. Thus, for instance, Euphranor suggests
that although there is no idea either of force in general or of particular forces,
nevertheless,

there are very evident propositions or theorems relating to force,

which contain useful truths. . . And if, by considering this doctrine of
force, men arrive at the knowledge of many inventions in Mechanics,
and are taught to frame engines, by means of which things difficult
and otherwise impossible may be performed. . . shall we deny that
[the doctrine of force] is of use either in practice or speculation,
because we have no distinct idea of force? (Alc, 7,7:295–296)
As Berkeley had emphasized in De Motu, the rules for the word ‘force’ are precise
mathematical rules that allow for predictive calculations that achieve practical
ends (DM, 36–39:40–41). Euphranor indicates that this sort of mathematical
15. For defense of the publicity of language in Berkeley, see Pearce 2017c, 37–41, 58–60,
76–78. On the consequences of this view for Berkeley’s metaphysics and epistemology, see
Pearce 2017a.
Berkeley’s Theory of Language 13

language can be used as a model to understand language in general: “what is

true of algebraic signs, is also true of words or language, modern algebra being
in fact a more short, apposite, and artificial sort of language” (Alc, 7,14:307).
Euphranor draws from this discussion the following conclusions:
Thus much, upon the whole, may be said of all signs: that they
do not always suggest ideas signified to the mind: that when they
suggest ideas, they are not general abstract ideas: that they have
other uses besides barely standing for and exhibiting ideas, such as
raising proper emotions, producing certain dispositions or habits of
mind, and directing our actions in pursuit of that happiness, which
is the ultimate end and design, the primary spring and motive that
sets rational agents at work: [that signs may imply or suggest the
relations of things; which relations habitudes or proportions, as they
cannot be understood but by the help of signs, so being thereby
expressed and confuted, they direct and enable us to act with regard
to things;]16 that the true end of speech, reason, science, faith, assent
in all its different degrees, is not merely, or principally, or always the
imparting or acquiring of ideas, but rather something of an active
operative nature, tending to a conceived good; which may sometimes
be obtained, not only although the ideas marked are not offered to
the mind, but even although there should be no possibility of offering
or exhibiting any such idea to the mind (Alc, 7,14:307).

This passage creates serious problems for emotivist interpretations of Al-

ciphron, since Euphranor here makes a claim about “speech, reason, science,
faith, assent in all its different degrees.” He is clearly not talking about “the
periphery of language” as Bennett (1971, 54) claims.
It was this observation that led Antony Flew to argue that Berkeley was here
endorsing a theory of ‘meaning as use’ that anticipated the later Wittgenstein
(Flew [1974] 1993).17 According to the particular version of this account that
I have developed in detail (Pearce 2017c), Berkeley holds that language is to
be analyzed by specifying for each utterance the conventional rules followed by
speaker and hearer, and the ends at which this discursive practice aims. This is
just what Berkeley does in the example of force above: the word ‘force’ is used
according to rules given by the Newtonian laws and the vector calculus and this
use aims at predictions which allow us “to frame engines, by means of which
things difficult and otherwise impossible may be performed” (Alc, 7,7:295).
On this interpretation, the communication of ideas is just one of the many
ways words accomplish their operative ends. In this respect the use theory is
similar to the speaker intentions theory: both see the communication of ideas
as a special case that fits within a more general theory of language. This aligns
with Berkeley’s remark in the published Introduction that the communication
of ideas marked by words is “sometimes entirely omitted, when these [operative
16. Bracketed text added in 1752.
17. Another early version of the use theory interpretation can be found in Woozley 1976.
Berkeley’s Theory of Language 14

ends] can be obtained without it, as I think not infrequently happens in the
familiar use of language” (PHK, Intro 20:37).
The use theory also provides a plausible interpretation of Berkeley’s remarks
on operative language more broadly. Like the emotivist interpretation of oper-
ative language, the use theory interpretation takes seriously Berkeley’s claims
that “passions. . . arise immediately in the mind upon the perception of certain
words” (PHK, Intro 20:37). However, like the speaker intentions theory, the use
theory also takes seriously Berkeley’s insistence that religious language is not
a special case but gets to be meaningful in the same way as ordinary bits of
language, such as number words (Alc, 7,5:293).
The use theory is also able to explain how Berkeley is able to make sense of
belief without mental propositions. In Alciphron, Berkeley claims to be giving
an account of “assent [i.e., in our terms, belief] in all its different degrees” (Alc,
7,14:307). Since, as I argued in §2, Berkeley rejects the mental proposition, the
object of assent must be a verbal proposition, i.e., a sentence. To assent to a
sentence is to regard it as true.
According to Euphranor, “Science [i.e., scientific knowledge] and faith agree
in this, that they both imply an assent of the mind” (Alc, 7,11:303). The
character Crito (who typically also speaks for Berkeley in Alciphron) insists,
however, that:
faith . . . is . . . placed in the will and affections rather than in the un-
derstanding, and producing holy lives, rather than subtle theories.
Faith, I say, is not an indolent perception, but an operative persua-
sion of the mind, which ever works some suitable action, disposition,
or emotion in those who have it. (Alc, 7,10:301)
According to the use theory, this is not a point of contrast between faith and
scientific knowledge, but a point of commonality: all assent is a matter of being
disposed to relate to the world around one in a certain way, which is practical
and emotional as well as cognitive.18
The most serious problem for such an interpretation, however, is how assent
(belief), understood on this model, could be a kind of taking for true. Admit-
tedly, Berkeley never directly addresses this issue, and any reconstruction of his
views on truth must be somewhat speculative. I have attempted such a specula-
tive reconstruction elsewhere (Pearce 2017c, 157–171).19 It suffices here to note
that since, as I argued in §2, Berkeley rejects the mental proposition and there-
fore could not inherit his predecessors’ conception of truth, every interpretation
of Berkeley is faced with this difficulty.

4.2 Meaning as Use in the Manuscript Introduction

When Flew originally proposed his proto-Wittgensteinian interpretation of Al-
ciphron VII, he suggested that this represented a fundamental departure from
18. For a more detailed exposition and defense of this general account of Berkeley on assent,
see Pearce 2017c, 139–157.
19. For alternative reconstructions, see Olscamp 1970, ch. 5; Schwartz 2010, §4.2.2.
Berkeley’s Theory of Language 15

Berkeley’s views in the early works (Flew [1974] 1993, 214). However, John
Russell Roberts (2007, ch. 2; 2017) has argued that Berkeley had this theory in
mind, at least in its basic outline, when he wrote the Manuscript Introduction
in 1708. I agree.
In both the manuscript and published versions of the Introduction, Berkeley
describes his task as “to try if I can discover what those Principles are, which
have introduc’d all that Doubtfulness & Uncertainty, those Absurdities & Con-
tradictions into the several Sects of Philosophy” (MI, 4:65, crossed out text
omitted; compare PHK, Intro 4:26). In the manuscript Berkeley goes on, al-
most immediately, to identify the culprit as “the opinion that there are Abstract
Ideas or General Conceptions of Things” (MI, 6:67). In this text, language is
not introduced until §16, where Berkeley includes two lengthy quotations from
Locke (E, 2.11.10–11:159–160) to the effect that non-human animals’ lack of
language is evidence that they lack abstract ideas. Berkeley uses these texts
to emphasize the tight connection between abstraction and a particular the-
ory of language (MI, 17–18:77–81). §6 of the manuscript has been extensively
marked up, indicating Berkeley’s dissatisfaction with his original formulation.
In the published version, Berkeley has decided to emphasize the connection
to language from the very beginning: before even mentioning abstract ideas,
he introduces the project of the Introduction by writing, “In order to prepare
the mind of the reader for the easier conceiving what follows, it is proper to
premise somewhat, by way of introduction, concerning the nature and abuse of
language” (PHK, Intro 6:27).
This does not represent a change in Berkeley’s position or his approach to
the issues. Rather, it provides additional emphasis to prevent the reader from
missing the point. In both the manuscript and published versions, immediately
after completing the case against abstraction, Berkeley identifies a mistaken
theory of language (namely, the ideational theory) as the source of the doctrine
of abstraction (MI, 30–33:95–101; PHK, Intro 18–19:36–37). He goes on to
argue, again in both versions, that there are other words that are even farther
divorced from ideas than general names for sensible objects.
Additionally, while Berkeley does not present his theory in the same explicit
and detailed fashion as he does in Alciphron, the particular examples he dis-
cusses conform to his later theory.20 In his discussions of ‘triangle’ and ‘animal’,
Berkeley treats the definition as a rule for determining which objects ‘have a
right to be called by’ a particular name (MI, 32–34:99–103). In the discussion
of ‘good things’ and ‘reward’, Berkeley discusses how we learn to give a cer-
tain practical and emotional response to these words, and how that response
accomplishes various practical goods (MI, 36–37:105–109). In the discussion of
‘Aristotle’, Berkeley speaks of the “close and immediate . . . Connexion . . . long
Custom [has] established betwixt the very Word Aristotle and the motions of
Assent and Reverence in the Minds of some Men” (MI, 38:111). The point of the
discussion is to show how the learned rule connecting the name ‘Aristotle’ (NB:
the name, not the philosopher) to these mental responses is used by the ‘School-
20. For a more detailed defense of this claim, see Pearce 2017c, ch. 2.
Berkeley’s Theory of Language 16

man’ to accomplish a certain end (“to dispose you to receive his Opinion with
. . . deference and Submission”). This rule is established by ‘custom’. Finally,
the discussion of ‘virtue’ and other terms of moral praise and blame is focused
on explaining how the moral emotions attached to these words by conventional
rules serve to motivate actions (MI, 41–42:111–112). This also neatly explains
why Berkeley here speaks not only of the intentions of a particular speaker on a
particular occasion (as Williford emphasizes), but also of ‘the ends of language’
(MI, 40:111), that is, the ends at which this conventional practice aims. Thus
it appears that, already in the Manuscript Introduction, Berkeley’s view is that
the way to understand any particular use of words is to understand the practical
ends at which that use aims and the conventional (‘customary’) rules by means
of which those ends are accomplished.
Admittedly, there are comments in both the manuscript and published ver-
sions that appear to favor the modified ideational theory over the use theory.
The strongest of these statements is found in MI, 30:97:
when any Truth whether general or particular is made known to me
by words So that I rightly apprehend the Ideas contained in it, I
see no manner of reason, why I may not omit the Words, and yet
retain as full and clear a Conception of the Ideas themselves, as I
had while they were cloathed with Words. Words being, so far as
I can see, of use [only]21 for Recording and Communicating, but
not absolutely apprehending Ideas. [Crossed out, bracketed, and
inserted text omitted.]
This text, as it stands in the first stratum, in fact appears to endorse the
unmodified ideational theory, but this certainly cannot be Berkeley’s view. Ac-
cordingly, it is unsurprising that this text is extensively marked up and then
entirely omitted from the published version. Most importantly, Berkeley modi-
fied this text to restrict the scope of his claim to “any Truth whether about gen-
eral or particular Ideas.” (The italicized words are inserted above caret marks.)
Further, in its context, this text was likely intended as ad hominem to begin
with: Berkeley has just noted that the abstractionists often claim that words
are necessary for knowledge of general truths, but this, he alleges, is inconsistent
with the ideational theory of meaning which the abstractionists endorse.
Another text in the Introduction that appears to support the modified
ideational theory is this one from the conclusion:
Since therefore words are so apt to impose on the understanding,
whatever ideas I consider, I shall endeavour to take them bare and
naked into my view, keeping out of my thoughts, so far as I am able,
those names which long and constant use has so strictly united with
them. (PHK, Intro 21:38–39; compare MI, 47:115)
This, Berkeley says, is necessary in order “to clear the first principles of knowl-
edge from the embarrass and delusion of words” (PHK, Intro 25:40).
21. Inserted above a caret.
Berkeley’s Theory of Language 17

However, these passages can be made to support the modified ideational

theory only by removing from their context. In fact, in these same concluding
pages of the Introduction, Berkeley blames the ideational theory of meaning and
the associated doctrine of abstraction for the problematic conflation of words
and ideas (PHK, Intro 23:39–40; cf. MI, 58–60:122–123). Further, he writes,
“He that knows he has no other than particular ideas, will not trouble himself
in vain to find out and conceive the abstract idea annexed to any name. And he
that knows names do not always stand for ideas, will spare himself that labour,
of looking for ideas, where there are none to be had” (PHK, Intro 24:40; cf.
MI, 61:123–135). In other words, Berkeley’s view is emphatically not that in
order to ensure that there is a ‘cognitive’ meaning attached to our words, or in
order to find the truth, we need to use only words that stand for ideas and focus
our attention on the ideas rather than the words. Instead, Berkeley holds, we
need to recognize that the relationship between words and ideas is much more
complex than the ideational theory would have us believe. Having recognized
this fact, we will be able to separate words from ideas, and this will save us
from a variety of philosophical errors (Pearce 2017c, 59–60).

4.3 Visual Language and the Use Theory

In the 1709 Essay toward a New Theory of Vision, Berkeley argues that “the
proper objects of vision constitute an universal language of the Author of na-
ture” (NTV, 147:231). According to this theory, which is also defended in the
fourth dialogue of Alciphron,
the proper objects of sight [which] are light and colours, with their
several shades and degrees. . . suggest and mark out to us the dis-
tances, figures, situations, dimensions, and various qualities of tan-
gible objects: not by similitude, nor yet by inference of necessary
connexion, but by the arbitrary imposition of Providence: just as
words suggest the things signified by them. (Alc, 4,10:154)
This text and others like it are naturally read as holding that the ‘words’ of the
visual language serve primarily to signify ideas, and this might be thought to
support the modified ideational theory as against the use theory.22
My response to this objection is twofold: first, Berkeley frequently empha-
sizes the operative character of the visual language, just as he frequently empha-
sizes the operative character of human languages. Second, the visual language
in fact does not—and, given Berkeley’s immaterialism, cannot—have the kind
of external signification attributed to language by the ideational theory.
The first point is simple enough. In the Essay toward a New Theory of
Vision, Berkeley states his conclusion as follows:
the proper objects of vision constitute an universal language of the
Author of nature, whereby we are instructed how to regulate our
22. I thank Samuel Rickless and Keota Fields for pressing this objection.
Berkeley’s Theory of Language 18

actions in order to attain those things that are necessary to the

preservation and well-being of our bodies, as also to avoid whatever
may be hurtful and destructive of them. It is by their information
that we are principally guided in all the transactions and concerns
of life. (NTV, 147:231)
Similarly, in discussing the interpretation of nature in the Principles, Berkeley
writes:
As in reading other books, a wise man will choose to fix his thoughts
on the sense and apply it to use. . . so in perusing the volume of
nature. . . We should propose to our selves. . . to recreate and exalt
the mind, with a prospect of the beauty, order, extent, and variety
of natural things; hence by proper inferences to enlarge our notions of
the grandeur, wisdom, and beneficence of the Creator; and lastly, to
make the several parts of the Creation, so far as in us lies, subservient
to the ends they were designed for, God’s glory, and the sustentation
and comfort of our selves and fellow-creatures. (PHK, 109:89)
Finally, in Alciphron IV, we read:
while [the visual language] informs, it amuses and entertains the
mind with a singular pleasure and delight. . . it answers so apposite
to the uses and necessities of mankind, informing us more distinctly
of those objects whose nearness and magnitude qualify them to be
of greatest detriment or benefit to our bodies. (Alc, 4,15:160)
When the visual language ‘instructs’, it is not conveying Lockean mental
propositions but rather instructing us what to do. Even when it is said to
‘inform’ this information is subordinate to operative ends. Indeed, if the use
theory is correct then informing (i.e., helping us to form correct beliefs) cannot
be separated from instructing (i.e., telling us what to do): a belief, on the use
theory, is just a complex of emotional, cognitive, and behavioral dispositions.
The system of visual signs is, according to Berkeley, above all a practical system
aimed at human flourishing. As Berkeley puts it in Siris, the regular course
of nature is “necessary to assist, not the Governor, but the governed” (Siris,
160:85).
In Berkeley’s treatment of the visual language, as in his treatment of human
languages, the suggestion of ideas is subordinated to practical ends. This, how-
ever, is not by itself inconsistent with the ideational theory: proponents of that
theory often said that communication of ideas was needed by humans because
of our practical need to live in society (see E, 3.1.1:402). The distinctive claim
accepted by the ideational theory and rejected by the use theory is that a word
gets to be meaningful by signifying some one particular idea, which in turn
represents the external object. Does Berkeley’s account of the visual language
function in this way?
In the New Theory of Vision, Berkeley does frequently write as though visual
ideas signify tangible ideas which in turn represent mind-independent tangible
Berkeley’s Theory of Language 19

objects. As is well-known, in the Principles Berkeley admits that in the New

Theory he had assumed that tangible objects exist outside the mind. Correcting
this “vulgar error,” Berkeley writes:
in strict truth the ideas of sight, when we apprehend by them dis-
tance and things placed at a distance, do not suggest or mark out to
us things actually existing at a distance, but only admonish us what
ideas of touch will be imprinted in our minds at such and such dis-
tances of time, and in consequence of such and such actions. (PHK,
44:59, emphasis added)
In ‘suggesting’ tangible ideas, Berkeley here tells us, visual ideas do not achieve
reference to any external objects. Instead, they provide practical ‘admonitions’
by forming correct action-guiding expectations about what ideas of touch we
might have in the future if we perform certain actions.
In fact, writers attempting to apply an ideational account of meaning to the
visual language have often been puzzled regarding the interaction between the
theory of visual language and immaterialism (see, e..g, Mabbott 1931, 26-27;
Olscamp 1970, 32-33; Winkler 1989, 21). In the Principles, Berkeley speaks not
merely of interpreting visual signs but of “perusing the volume of nature” (PHK,
109:89). In Siris, he describes “the phenomena of nature” as “a most coherent,
entertaining, and instructive Discourse” (Siris, 254:121), and in a passage added
to the 1752 edition of Alciphron he explicitly affirms that all sensory ideas are
signs, and not only visual ideas (Alc, 4,12:156–157), though he here denies that
smell and taste are sufficiently complex to count as language. Thus the tangible
objects signified by visual ideas are themselves only more signs.
If Berkeley applied the ideational theory to the visual language, then we
would expect visual ideas to signify some other ideas which in turn represented
something extra-linguistic. But this is not what we find, at least for broad
swathes of the visual language. For the most part, visual signs simply pro-
vide instructions on how to navigate the sign system itself. Like the sentence
‘Melampus is an animal’ which, according to the Manuscript Introduction, sim-
ply lets us know that the name ‘animal’ is rightly applied to Melampus, our
visual stimulus, for the most part, serves simply to indicate the next move in
a language-game. The ‘suggestion’ or ‘signification’ of which Berkeley so often
speaks is most often a suggestion only of other signs.23

4.4 Reference in the Use Theory

In the visual language, as in human languages (according to the use theory),
most words lack extra-linguistic reference.24 However, it will be objected, Berke-
23. For a detailed account of the visual language in light of the use theory, see Pearce 2017c,
ch. 9.
24. What I, in Pearce 2017c, call ‘quasi-referring terms’ do not have extra-linguistic reference,
since the referents of such words (e.g., physical forces) are created by the conventions governing
the use of their names.
Berkeley’s Theory of Language 20

ley surely cannot hold that all words lack extra-linguistic reference. Spirits and
ideas, at least, must surely exist prior to language.
The use theory can, however, accommodate this. Even Wittgenstein does not
deny that naming is one of the things we do with words (see, e.g., Wittgenstein
1953, §1.7). What use theorists reject is the view that naming or referring has
some kind of privileged status among the many things we do with words (see
Wittgenstein 1953, §§1.11–14).
The use theory, then, has no problem admitting that I can use words like
‘red’ to name my ideas, and that I can use words like ‘I’ and ‘myself’ to name my
own soul or spirit of which I have immediate consciousness (see DHP, 231).25
Admittedly, there is some unclarity regarding exactly how Berkeley believes he
can secure reference to other minds and the ideas within them. However, as is
widely recognized,26 this is a difficult issue on any interpretation.

4.5 Historical Continuity and the Use Theory

A further objection to the use theory is based on a methodological principle we
might call ‘the principle of historical continuity’. Contrary to a myth perpet-
uated by some old-fashioned histories, philosophical positions and arguments
do not arise ex nihilo from the minds of a select few geniuses. Instead, they
emerge in the course of philosophical debates and discussions and are shaped
by historical developments both internal and external to the practice of phi-
losophy. Thus the views of even the most original philosophers must be seen
as developing out of or responding to something in their cultural environment,
and we should approach with suspicion any interpretation that attributes to a
philosopher a “revolutionary and historically premature insight,” such as the
one Antony Flew ([1974] 1993, 216) attributed to Berkeley on the subject of
language. Although Berkeley’s remarks on language are clearly highly original,
and there is even some evidence that Berkeley himself regarded them as ‘revolu-
tionary’, a plausible interpretation ought to make Berkeley’s strategy intelligible
in its historical context, and not read the 20th century back into it, as the use
theory interpretation appears to do.
In fact, however, it is intelligible in Berkeley’s context that he should have
developed a use theory of language. The Principles and Alciphron—the two
published works in which Berkeley presents his theory of language—both clearly
position themselves within the deist controversy of the 18th century. In Alci-
phron, Berkeley’s argument for his theory of language is framed as a response
to John Toland’s deistic argument for the claim that religious ‘mysteries’, such
as the Trinity, were (as Berkeley’s character Alciphron puts it), “empty notions,
or, to speak more properly. . . mere forms of speech, which mean nothing, and
are of no use to mankind” (Alc, 7,1:287; see Toland 1696).
25. Wittgenstein would, of course, disagree with Berkeley regarding our knowledge of the
self and its states. The use theory interpretation of Berkeley should certainly not be seen as
assimilating Berkeley to Wittgenstein in all respects.
26. See, e.g., Adams 1973; Cummins 1982; Flage 1985; Bettcher 2007; Pearce 2017c, ch. 7;
2018.
Berkeley’s Theory of Language 21

Toland saw the mysteries of religion as a tool of tyranny: the clergy, backed
by the state, seek to undermine the intellectual autonomy of the laity by forcing
them to repeat whatever nonsense words the church prescribes.27 Many other
deistic writers, including Matthew Tindal (1706, 20–23; [1730] 1732, 27–30, 58–
60, 93–99, 213–217, 331)28 and Anthony Collins (1713, 112–122),29 argued that
the attempt to impose doctrines on the laity was contrary to the stated justifica-
tion for government support of religion, namely the state’s interest in promoting
virtue among its citizens.30 Although many Anglican clergy, including Berke-
ley,31 supported some form of religious toleration, any defense of Anglicanism
must necessarily involve a defense of the mysteries as obligatory articles of faith
within a state-supported church.32 If this promotion of the mysteries is to be
justified as part of the established church’s activity, it must play some role in
the church’s promotion of virtue, and this is precisely what many Anglican writ-
ers, including Berkeley, argue (see, e.g., Stillingfleet 1691, 28; 1697, lvii; Astell
[1705] 2013, §64; Alc, 7,8–10:296–303). Furthermore, even before Berkeley, this
argument about the (alleged) salutary moral effects of belief in, e.g., the Trinity
is often tangled up with the argument about the meaningfulness of the words
used to express such belief. This entanglement can be observed in a particularly
strong form in a philosopher and clergyman whose influence on Berkeley is well
documented: William King, the Archbishop of Dublin (see King 1709).33
In Berkeley’s context, the debate about the meaningfulness of Christian
mysteries and the debate about the tendency of belief in those mysteries to
promote virtue were already entwined. Explicitly in Alciphron, and implicitly
27. A very strong statement of this view, which may have influenced Toland, can be found
in the opening of the anonymous Account of the Word Mystery:
MYSTERY and Persecution are the Tutelar God and Goddess of the new Sys-
tems, framed by Worldly Christians. Upon those two Antichristian Principles
they ground all their Arguments: Mystery makes up the Premises, and Perse-
cution the Consequence or Conclusion. (Anonymous 1691, 3)
Also see Berkeley’s comments on this view in Alc, 1,6–7:40–42.
28. Berkeley discusses the central argument of Tindal (1730) 1732 (referring to Tindal by
the pseudonym ‘Glaucus’) in Alc, 6,18–19:254–255, and mentions Tindal 1706 in TVV, 2:251,
5:253.
29. Alciphron discusses Collins under the pseudonym ‘Diagoras’ (Alc, 1,12:52, 4,16–22:163–
171), and Berkeley makes explicit reference to Collins in TVV, 6:254. On Berkeley’s inter-
pretation of, and response to, Collins, see O’Higgins 1976; Berman 1990, ch. 3; Taranto 2010;
Pearce 2018.
30. In addition to Tindal and Collins, see Toland 1720, 133–136.
31. It is unclear exactly what Berkeley takes to be the correct boundaries of religious tol-
eration, and there is some evidence that he becomes more liberal about this later in his
career. However, both early and late, he clearly supports the legal toleration of at least some
non-Anglican Christians (RGB, 70–71; Q, 191:120–121, 256–257:126, 1,297–299:160).
32. Tindal (1706, 22) argues that it is inconsistent to support religious toleration while also
supporting a state church with obligatory doctrines: “if it be the highest Injustice to force
Men to profess such Speculative Opinions as they don’t believe, it can’t savour much of Justice
to make ’em contribute [through taxation] to the Support of [those opinions].”
33. Berkeley explicitly references King’s sermon at CGB, 36, and likely also has King in mind
at Alc, 4,16–18:163–166. Additionally, I have argued that King’s sermon forms an important
part of the background to DHP (Pearce 2018). On the pragmatist tendencies of King’s sermon,
see Berman 1976.
Berkeley’s Theory of Language 22

in the manuscript and published introductions, Berkeley is responding to these

debates. He responds to the meaningfulness debate by arguing that the tendency
of forms of words to affect the life and character of those who employ them is
constitutive of their meaningfulness. Contrary to the emotivist interpretation,
Berkeley adopts this account not only for certain special cases but for language
in general. This is indeed a radical and highly original move, but it is also
intelligible as a response to an existing debate, as the principle of historical
continuity requires. Berkeley’s view is that not only a special class of religious
words, but all words get to be meaningful by virtue of the role they play in
shaping our engagement with the world around us, including especially our
relations to our fellow human beings. Words are tools for promoting human
flourishing. Sometimes words perform their function by suggesting ideas, and
sometimes they perform it without suggesting ideas.

5 Conclusion
According to the ideational theory of meaning, humans have certain ideas to-
gether with a capacity to perform acts of judgment assembling these ideas
into truth-evaluable mental states with subject-predicate structure. Individual
words, on this view, typically signify the speaker’s ideas, while complete sen-
tences signify these truth-evaluable mental states (mental propositions). Both
early and late in his career, Berkeley explicitly attacks this view. I have argued
that Berkeley’s attack is more radical than most interpreters have recognized.
Berkeley in fact entirely rejects the notion that meaningful words must signify
what takes place in the mind of the speaker. Instead, both early and late in
his career, Berkeley holds that words get to be meaningful when they are used
according to rules in a conventional social practice aiming at some practical
goods.34

Abbreviations
Alc Berkeley, George. (1732) 1948–1957. Alciphron: or, the Minute Philoso-
pher . In BW, vol. 3.
BW Berkeley, George. 1948–1957. The Works of George Berkeley, Bishop of
Cloyne. Edited by A. A. Luce and T. E. Jessop. 9 vols. London: Thomas
Nelson and Sons.
CGB Hight, Marc A., ed. 2013. The Correspondence of George Berkeley. Cam-
bridge: Cambridge University Press.
34. A shortened version of this paper was presented at the University of Wisconsin Milwaukee
Berkeley Workshop, 20 October 2018. I thank all of the workshop participants, and especially
my commentator Keota Fields, for helpful discussion of these issues. Additionally, I am
grateful to Samuel Rickless and Shelley Weinberg for helpful comments previous drafts.
Berkeley’s Theory of Language 23

CSM Cottingham, John, Robert Stoothoff, Dugald Murdoch, and Anthony

Kenny, trans. 1984–1991. The Philosophical Writings of Descartes. 3 vols.
Cambridge: Cambridge University Press.
DHP Berkeley, George. (1713) 1948–1957. Three Dialogues Between Hylas and
Philonous. In BW, vol. 2.
DM Berkeley, George. (1721b) 1948–1957. An Essay on Motion. In BW,
4:243–267. Latin edition: De Motu: Sive, De Motus Principio & Natura,
et de Causa Communicationis Motuum. In BW, vol. 4.
E Locke, John. (1690) 1975. An Essay Concerning Human Understanding.
Edited by Peter H. Nidditch. Oxford: Oxford University Press.
MI Berkeley, George. 1987. George Berkeley’s Manuscript Introduction: An
Editio Diplomatica. Edited by Bertil Belfrage. Oxford: Doxa.
NTV Berkeley, George. (1709) 1948–1957. An Essay Toward a New Theory of
Vision. In BW, vol. 1.
PC Berkeley, George. 1989. Philosophical Commentaries. Edited by George
Hasson Thomas. The Philosophy of George Berkeley. Ohio: Garland.
PHK Berkeley, George. (1710) 1948–1957. A Treatise Concerning the Princi-
ples of Human Knowledge. In BW, vol. 2.
Q Berkeley, George. (1735) 1948–1957. The Querist: Containing Several
Queries Proposed to the Consideration of the Public. In BW, 6:105–181.
RGB Berkeley, George. (1721c) 1948–1957. An Essay Towards Preventing the
Ruin of Great Britain. In BW, 6:68–87.
Siris Berkeley, George. (1744) 1948–1957. Siris: A Chain of Philosophical Re-
flexions and Inquiries Concerning the Virtues of Tar-water, and divers
other Subjects connected together and arising One from Another . In BW,
5:31–164.
TVV Berkeley, George. (1733) 1948–1957. The Theory of Vision, or Visual
Language, Shewing The Immediate Presence and Providence of a Deity,
Vindicated and Explained . In BW, vol. 1.

References
Adams, Robert Merrihew. 1973. “Berkeley’s ‘Notion’ of Spiritual Substance.”
Archiv für Geschichte Der Philosophie 55 (1): 47–69.
Anonymous. 1691. An Impartial Account Of the Word Mystery, As it is taken
in the holy Scripture. London.
Arnauld, Antoine, and Claude Lancelot. (1660) 1975. General and Rational
Grammar: The Port-Royal Grammar. Edited and translated by Jacques
Rieux and Bernard E. Rollin. Janua Linguarum. The Hague: Mouton.
Berkeley’s Theory of Language 24

Arnauld, Antoine, and Pierre Nicole. (1662) 1996. Logic or the Art of Think-
ing: Containing besides common rules, several new observations appropri-
ate for forming judgments. Edited and translated by Jill Vance Buroker.
Cambridge Texts in the History of Philosophy. Cambridge: Cambridge Uni-
versity Press.
Ashworth, E. J. 1981. “‘Do Words Signify Ideas or Things?’: The Scholastic
Sources of Locke’s Theory of Language.” Journal of the History of Philos-
ophy 19 (3): 299–326.
. (1984) 1998. “Locke on Language.” In Chappell 1998, 175–198.
Astell, Mary. (1705) 2013. The Christian Religion, as Professed by a Daughter
of the Church of England. Edited by Jacqueline Broad. The Other Voice in
Early Modern Europe: Toronto Series 24. Toronto: Iter Inc. / Center for
Reformation and Renaissance Studies.
Ayer, Alfred Jules. (1946) 1952. Language, Truth, and Logic. 2nd ed. Mineola,
NY: Dover.
Belfrage, Bertil. 1986a. “Berkeley’s Theory of Emotive Meaning (1708).” History
of European Ideas 7 (6): 643–649.
. 1986b. “Development of Berkeley’s Early Theory of Meaning.” Revue
Philosophique de la France et de l’Etranger 176 (3): 319–330.
. 1987a. Editor’s commentary on George Berkeley’s Manuscript Introduc-
tion: An Editio Diplomatica, by George Berkeley, edited by Bertil Belfrage.
Oxford: Doxa.
. 1987b. Introduction to George Berkeley’s Manuscript Introduction: An
Editio Diplomatica, by George Berkeley, edited by Bertil Belfrage. Oxford:
Doxa.
Bennett, Jonathan Francis. 1971. Locke, Berkeley, Hume: Central Themes. Ox-
ford: Clarendon Press.
Berman, David. 1976. Introduction to Archbishop King’s Sermon on Predesti-
nation, by William King. Edited by Andrew Carpenter. Irish Writings from
the Age of Swift 4. Dublin: Cadenus Press.
. 1990. A History of Atheism in Britain: From Hobbes to Russell. London:
Routledge.
. 1994. George Berkeley: Idealism and the Man. Oxford: Clarendon Press.
Bettcher, Talia Mae. 2007. Berkeley’s Philosophy of Spirit: Consciousness, On-
tology, and the Elusive Subject. London: Continuum.
Bordner, S. Seth. 2017. “If We Stop Thinking About Berkeley’s Problem of
Continuity, Will It Still Exist?” Journal of the History of Philosophy 55
(2): 237–260.
Berkeley’s Theory of Language 25

Buroker, Jill Vance. 1996. “Arnauld on Judging and the Will.” In Interpreting
Arnauld, edited by Elmar J. Kremer, 3–12. Toronto: University of Toronto
Press.
Chappell, Vere, ed. 1998. Locke. Oxford Readings in Philosophy. Oxford: Oxford
University Press.
Collins, Anthony. 1713. A Discourse of Free-Thinking: Occasion’d by the Rise
and Growth of a Sect Call’d Free-Thinkers. London.
Cummins, Phillip D. 1982. “Hylas’ Parity Argument.” In Berkeley: Critical and
Interpretive Essays, edited by Colin Murray Turbayne, 283–294. Minneapo-
lis: Minnesota University Press.
Flage, Daniel E. 1985. “Berkeley’s Notions.” Philosophy and Phenomenological
Research 45 (3): 407–425.
Flew, Antony. (1974) 1993. “Was Berkeley a Precursor of Wittgenstein?” In
George Berkeley: Alciphron, or the Minute Philosopher in Focus, edited by
David Berman, 214–226. London and New York: Routledge.
Frankel, Melissa. 2009. “Something-We-Know-Not-What, Something-We-Know-
Not-Why: Berkeley, Meaning and Minds.” Philosophia 37 (3): 381–402.
Guyer, Paul. 1994. “Locke’s Philosophy of Language.” In The Cambridge Com-
panion to Locke, edited by Vere Chappell, 115–145. Cambridge: Cambridge
University Press.
Hacking, Ian. 1974. Why Does Language Matter to Philosophy? Cambridge:
Cambridge University Press.
Jakapi, Roomet. 2002. “Emotive Meaning and Christian Mysteries in Berkeley’s
Alciphron.” British Journal for the History of Philosophy 10 (3): 401–411.
. 2007. “Christian Mysteries and Berkeley’s Alleged Non-Cognitivism.”
In Reexamining Berkeley’s Philosophy, edited by Stephen H. Daniel, 188–
198. Toronto: University of Toronto Press.
King, William. 1709. Divine Predestination and Fore-knowledg, consistent with
the Freedom of Man’s Will. Dublin and London: J. Baker.
Losonsky, Michael. 1994. “Locke on Meaning and Signification.” In Locke’s Phi-
losophy: Content and Context, edited by G. A. J. Rogers, 123–141. Oxford:
Clarendon Press.
. 2007. “Language, Meaning, and Mind in Locke’s Essay.” In The Cam-
bridge Companion to Locke’s “Essay Concerning Human Understanding”,
edited by Lex Newman, 286–312. Cambridge: Cambridge University Press.
Mabbott, J. D. 1931. “The Place of God in Berkeley’s Philosophy.” Journal of
Philosophical Studies 6 (21): 18–29.
Berkeley’s Theory of Language 26

Marušić, Jennifer Smalligan. 2014. “Propositions and Judgments in Locke and

Arnauld: A Monstrous and Unholy Union?” Journal of the History of Phi-
losophy 52 (2): 255–280.
Mattern, Ruth Marie. (1978) 1998. “Locke: “Our Knowledge, Which All Consists
in Propositions”.” In Chappell 1998, 226–241.
O’Higgins, James. 1976. “Browne and King, Collins and Berkeley: Agnosticism
or Anthropomorphism?” Journal of Theological Studies 27 (1): 88–112.
Olscamp, Paul J. 1970. The Moral Philosophy of George Berkeley. The Hague:
Martinus Nijhoff.
Ott, Walter. 2002. “Propositional Attitudes in Modern Philosophy.” Dialogue
41 (3): 551–568.
Pearce, Kenneth L. 2015. “Port-Royal.” In Routledge Encyclopedia of Philos-
ophy, edited by Tim Crane. London: Routledge. https : / / www . rep .
routledge.com/articles/port-royal/v-2.
. 2017a. “Berkeley on Unperceived Objects and the Publicity of Lan-
guage.” History of Philosophy Quarterly 34 (3): 231–250.
. 2017b. “How Berkeley’s Gardener Knows his Cherry Tree.” Pacific
Philosophical Quarterly 98 (S1): 553–576.
. 2017c. Language and the Structure of Berkeley’s World. Oxford: Oxford
University Press.
. 2018. “Matter, God, and Nonsense: Berkeley’s Polemic against the Free-
thinkers in the Three Dialogues.” In Berkeley’s Three Dialogues: New Es-
says, edited by Stefan Storrie. Oxford: Oxford University Press.
Pritchard, Timothy. 2013. “Locke and the Primary Signification of Words: An
Approach to Word Meaning.” British Journal for the History of Philosophy
21 (3): 486–506.
Rickless, Samuel C. 2008. “Is Locke’s Theory of Knowledge Inconsistent?” Phi-
losophy and Phenomenological Research 77 (1): 83–104.
Roberts, John Russell. 2007. A Metaphysics for the Mob: The Philosophy of
George Berkeley. Oxford: Oxford University Press.
. 2017. “Berkeley on the Philosophy of Language.” In The Bloomsbury
Companion to Berkeley, edited by Richard Brook and Bertil Belfrage, 421–
434. London: Bloomsbury.
Schaar, Maria van der. 2008. “Locke and Arnauld on Judgment and Proposi-
tion.” History and Philosophy of Logic 29 (4): 327–341.
Berkeley’s Theory of Language 27

Schwartz, Claire. 2010. “Berkeley and His Contemporaries: The Question of

Mathematical Formalism.” In George Berkeley: Religion and Science in the
Age of Enlightenment, edited by Silvia Parigi, 43–56. International Archives
of the History of Ideas 201. Dordrecht: Springer.
Stillingfleet, Edward. 1691. The Mysteris of the Christian Faith Asserted and
Vindicated: In a Sermon Preached at S. Laurence-Jewry in London, April
the 7th 1691. London: Henry Mortlock.
. 1697. A Discourse in Vindication of the Doctrine of the Trinity: With
An Answer to the Late Socinian Objections Against it from Scripture, An-
tiquity and Reason. 2nd ed. London: Henry Mortlock.
Taranto, Pascal. 2010. “Le personnage de Diagoras.” In Berkeley’s Alciphron:
English Text and Essays in Interpretation, edited by Laurent Jaffro, Geneviéve
Brykman, and Claire Schwartz, 361–370. Hildesheim: Georg Olms Verlag.
Tindal, Matthew. 1706. The Rights of the Christian Church Asserted: Against
the Romish, and all Other Priests who Claim an Independent Power Over
It. London.
. (1730) 1732. Christianity as Old as Creation: Or, The Gospel, a Repub-
lication of the Religion of Nature. 3rd ed. London.
Tipton, I. C. 1974. Berkeley: The Philosophy of Immaterialism. London: Methuen
& Co Ltd.
Toland, John. 1696. Christianity not Mysterious: Or, a Treatise Shewing, That
There is Nothing in the Gospel Contrary to Reason, Nor Above it: And that
no Christian Doctrine Can be Properly Call’d a Mystery. 2nd ed. London:
Sam. Buckley.
. 1720. Tetradymus. London.
Williford, Kenneth. 2003. “Berkeley’s Theory of Operative Language in the
Manuscript Introduction.” British Journal for the History of Philosophy 11
(2): 271–301.
Williford, Kenneth, and Roomet Jakapi. 2009. “Berkeley’s Theory of Meaning
in Alciphron VII.” British Journal for the History of Philosophy 17 (1):
99–118.
Winkler, Kenneth P. 1989. Berkeley: An Interpretation. Oxford: Clarendon
Press.
. 2009. “Signification, Intention, Projection.” Philosophia 37 (3): 477–
501.
Wittgenstein, Ludwig. 1953. Philosophical Investigations. Translated by G. E. M.
Anscombe. New York: Macmillan.
Woozley, A. D. 1976. “Berkeley’s Doctrine of Notions and Theory of Meaning.”
Journal of the History of Philosophy 14 (4): 427–434.