8

Locke and Reactions to Locke, 1700–1780

Nicholas Hudson

John Locke’s Essay Concerning Human Understanding ([1690]1975) inaugurated

a ‘linguistic turn’ in eighteenth-century thought, making the relationship
between language, thought, and the world a topic of energetic discussion and
debate in both British and continental philosophy (see Chapter 9). Yet Locke’s
formative influence on eighteenth-century language theory took a peculiar
form. It was less Locke’s achievement in making correct insights into the role
and function of language that inspired later writers than his perceived inaccu­
racies and mistakes. Although given credit, as the Abbé de Condillac wrote,
for being the first writer to handle the subject of language “like a philoso­
pher” ([1746]1971: 11), he was repeatedly accused of having either underrated
or overrated the mental role of language, using unclear or outdated termi­
nology, and a variety of other theoretical and methodological errors. Locke
has the paradoxical distinction of having launched a rich tradition of linguistic
study devoted in large part to his correction or even his rebuttal.
Disrespect for Locke’s linguistic thought has extended into our own time.
The main thesis of Locke’s theory of signification, “Words in their primary or
immediate Signification, stand for nothing, but the Ideas in the Mind of him that uses
them” (1975: iii, ii, §2, 405),1 has struck philosophers of language as “one of the
classic blunders in semantic theory” (Kretzmann 1977: 124). As John Stuart Mill
objected – when I say “the sun causes the day,” I certainly propose a descrip­
tion of objective reality rather than my mental experience ([1843]1851: i, 23–4).
Furthermore, whereas Locke tended to treat language solely as a means to
convey information, language has many other functions, and is meaningful
not by virtue of the speaker’s intentions but as the result of linguistic conven­
tions and the particular context of a speech act. As urged by Locke’s modern
apologists, however, such objections equally ignore the intellectual context
of his own discussion of language (Ashworth 1981, 1989). To assert that words

1 References to Locke 1975: iii = book 3, ii = chapter 2, §2 = section 2, 405 = page 405.

258

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

denote “ideas” rather than objective states of external reality was far from
new, but it aligned Locke with a rising tradition of philosophical thought that
challenged a powerful intellectual institution, the Aristotelian tradition of the
‘schools,’ that treated Latin syntax as a transparent mirror of external reality.
Locke wished to show that words do not automatically mirror an objec­
tive reality, and that many of the terms that fill scholastic philosophy, such as
‘substance,’ ‘essence,’ or ‘being,’ in fact refer to nothing in the world or even
in the minds of those who used this vocabulary. He wished, as he insisted,
to expose this “abuse” of language, and to set out clear criteria for enabling
meaningful philosophical discussion on topics that had been clouded by
empty verbiage (1975: iii, ix–x). His major criterion was that philosophers
must abandon any assumptions that could not be proved to derive from the
senses or reflection on the mind. His strict empiricism attempted to clear
philosophy not only of meaningless jargon, but also of belief in innate ideas
and in supposed faculties of the mind capable of directly apprehending real­
ity. The mind, as described by Locke, was originally comparable to “white
Paper” (1975: ii, i, 2, 104) imprinted with simple sensory data which became
more complex through further experience and education. The knowledge
built from these perceptions could be achieved without any innate faculties
except what he called “Intuition,” the capacity to recognize the agreement or
disagreement of ideas, memory or the capacity to store and recover ideas,
and what philosophers deceptively termed the “will,” really the basic drive to
pursue pleasure and avoid pain.
Locke’s theory of signification, that is, was really subsidiary to his primary
goal to make philosophical language more logical and certain. Nevertheless,
it is the great irony of Locke’s discussion of words that, far from banishing
the pernicious influence of words on philosophy, he found himself increas­
ingly obliged to allow words an evermore integral role in the evolution of
rational thought. In the empiricist tradition that he inspired, words became
‘constitutive’ of rational thought rather than being relegated to the role of
suspect signifiers for real things or ideas. We will also find, however, that this
purely empiricist model of understanding led increasingly to philosophical
impasses and insoluble paradoxes. These problems ultimately provoked the
reaffirmation of the belief, in a new form, that the human mind must indeed
possess special capacities beyond the senses and basic drives in order to create
language, which in turn constituted the very foundation of rational thought.
Beyond what he had foreseen, Locke launched a process of debate that ended
with the reaffirmation of some underlying structure to human language, and
by extension the creation of linguistic science as a distinct field.

259

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

Locke and the Scholastic Tradition

Locke would be frequently criticized for his continued use of a vocabulary
including terms such as ‘substance,’ ‘essence,’ ‘genera,’ and ‘species,’ bor­
rowed from the very tradition of philosophy that he was attacking. His use of
this terminology nonetheless clearly identifies the intended audience of the
Essay. From 1652 to 1665, Locke was a student and later a lecturer at Oxford
University, an institution still dominated by a tradition of philosophy derived
from medieval interpretations of Aristotle. It is indeed helpful to read the
Essay as an ongoing debate between an ardently Protestant and rationalist
young man and his Oxford tutors and colleagues who still adhered to the
doctrines of a largely Roman Catholic heritage resistant to the new philos­
ophy of Bacon, Descartes, and their seventeenth-century admirers. Locke
was required to master and then to teach logic books by seventeenth-century
authors whose names now seem insignificant – Blundeville (1599), Burgers­
dijck (1647), Smiglecius (1634), Sanderson (1618), du Trieu (1620). From the
eighteenth century onward, these ‘schoolmen’ have had the reputation of
slavishly rehearsing the doctrines of St. Thomas Aquinas and Duns Scotus.
While conventionality was indeed a feature of scholastic commentary, this
tradition had adjusted in various ways to ongoing challenges since the late
Middle Ages.
The first of these challenges, ‘nominalism,’ had an important influence on
Locke’s thinking. As most forcefully articulated by the fourteenth-century
philosopher William of Ockham, nominalism assailed the ‘moderate realism’
characteristic of standard medieval interpretations of Aristotle (see Chapter 5).
According to moderate realism, the relationship between substantives and
predicates in Latin grammar mirrored the relationship between ‘universals’
and ‘accidents’ in the real world. A ‘universal,’ signified by substantives such
as man, existed potentially (in potentia) as the ‘essence’ of each individual
being. This essence was realized as a metaphysical reality only by the human
intellect. The scholastic insistence that universals existed only potentially
in individual beings is important because of another challenge to scholastic
orthodoxy, Platonism, which maintained that universals existed apart from
the human intellect as Ideas. Platonic realism (as opposed to moderate real­
ism) nonetheless emerged only in the early sixteenth century after the recov­
ery of most of Plato’s oeuvre. It was thereafter usually noted, only to be listed
and rebutted in early modern logic texts. Ockham’s nominalism, on the con­
trary, insisted that universals were only individual mental images, or what
Ockham called ‘intentions.’ The individual intention was then extended as a

260

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

sign to designate all objects that bore a similarity to this mental image. Uni­
versals thus had no reality outside the human mind (see Carré 1946; Trent-
man 1970; Leff 1975; Adams 1987).
Although Locke advocated a new philosophy drawn from his enthusiastic
reading of Bacon, Descartes, Hobbes, and others, he inherited the problem­
atic terminology of ‘essence’ and ‘substance’ from the scholastics. He also
concurred with the virtual consensus in scholastic philosophy that the rela­
tionship between articulate words and mental concepts was merely conven­
tional or ‘arbitrary’ (ad placitum). Indeed, on this issue scholasticism and the
new philosophy were in agreement, for none of the contemporary thinkers
whom Locke admired denied that words were imposed by “voluntary Impo­
sition” (1975: iii, ii, §1, 405). Nevertheless, beginning with the rediscovery of
Platonic texts in the sixteenth century, some authors had ventured to specu­
late on the natural resemblance between words and things. In the early seven­
teenth century, moreover, interest in recovering the traces of some original
similarity between words and things in Adamic language (often thought to
be a purer version of Hebrew) became especially prevalent in Germany, as
exemplified by the influential mystic theology of Jacob Boehme (see Padley
1985–8: i, 86–8; Losonsky 2001: 105–19). As we will see, Locke’s general indif­
ference to these recent trends is ironic, for his own analysis of the ‘will’ denied
the possibility of totally ‘arbitrary’ choices, giving rise to renewed interest in
the ‘natural’ origins of words.
Locke claimed in the final version of the Essay that he had not originally
intended to spend much time discussing language. He had initially planned
only to refute belief in innate ideas, and to demonstrate how all knowledge
was built from ideas derived from the senses. “Upon a nearer approach,”
however, “I find, that there is so close a connection between Ideas and Words;
and all our abstract Ideas, and general Words, have so constant a relation
one to another, that it is impossible to speak clearly and distinctly of our
Knowledge, which all consists in Propositions, without considering, first, the
Nature, Use, and Signification of Language” (1975: ii, xxxiii, §19, 401; see also
iii, ix, §21, 488). Although this statement would be seized on by later critics
as showing that his consideration of language was belated and incomplete, it
is somewhat disingenuous. Early drafts of the Essay in 1671 show that he was
obsessed from the beginning with the scholastic confusion between words
and things ([1671]1936: 4–8). Nonetheless, Locke does seem to have been sur­
prised that he could not simply replace the scholastic conceptions of genera,
species, and essences with sensory ideas without repeatedly acknowledging
the role of words in the development of rational thought. ‘Development’ is

261

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

a key word here, for neither scholastic philosophers nor any previous phi­
losopher saw the need to trace the evolution of the understanding from its
origins in childhood. Locke, on the contrary, began with what he took to be
the building blocks of the understanding. These were ‘simple ideas,’ or the
child’s initial perceptions of color, tactile sensations, sounds, and other indi­
vidual sensations that cannot be defined, only experienced. Over time, the
mind would assemble these sensations into ‘complex ideas.’ To cite Locke’s
favorite example, the child might begin with the simple idea of a yellow color,
which, later joined with the tactile idea of a hard substance and with a series
of other ideas such as heaviness and fusibility, become the complex idea given
the English name ‘gold.’
In the empiricist tradition that the Essay inspired, philosophers found much
to question in this account of the origin of ideas. First, it seemed unlikely that
the child would build simple ideas up to complex ideas according to the scien­
tific method prescribed by Francis Bacon. Instead, the child would surely begin
with Locke’s complex ideas, such as the ideas named ‘nurse’ or ‘tree,’ which
would then be ‘analyzed’ into simple ideas. Second, Locke’s initial account of
the origin of the understanding fails to explain how the child would make either
a simple or complex idea into a ‘general idea.’ Locke’s goal had been to demon­
strate that the species and genera of scholastic philosophy, such as ‘man’ or
‘gold,’ were in fact merely collections of simple ideas. “Tis plain,” he wrote,
“that our distinct Species, are nothing but distinct complex Ideas, with distinct Names
annexed to them” (1975: iii, vi, §13, 448). But how did a particular abstract idea,
such as the man named ‘Peter,’ become the general idea ‘man’ denoted by that
name? On this question Locke found himself repeatedly compelled to admit
that the creation of abstract ideas seemed inseparable from the use of language.
“The sorting of Things by us, or the making of determinate Species,” he wrote,
was “in order to naming and comprehending them under general terms.” Peo­
ple sorted individual complex ideas “by certain obvious appearances, into Spe-
cies, that we may the easier, under general names, communicate our thoughts
about them” (1975: iii, vi, §30, 458). It was at this point that Locke took a signifi­
cant step beyond the nominalism of William of Ockham.

Locke’s Linguistic Turn

Having begun by refuting the linguistic errors of conventional Aristotelian-
ism, Locke had been led by his strict empiricism into giving a language a
far more integral function in the understanding than anticipated even by its
competing school, nominalism. Like Locke, William of Ockham had denied

262

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

the existence of metaphysical genera and species, arguing instead that an indi­
vidual “intention” came to designate a range of similar objects. Nonetheless,
so far from admitting that this generalization of a particular mental object
had anything to do with language, Ockham strictly separated words from
the mind. Moreover, Ockham did not question that genera and species truly
existed in nature, as identified by similarity, though he denied the existence
of metaphysical essences. Locke, by contrast, indicated that genera and spe­
cies by no means existed in reality, but were themselves “the Workmanship of
the Understanding” (1975: iii, iii, §13, 415), categories created for convenience
of language. The “Boundaries of the Species, are as Men, and not as Nature
makes them, if at least there are in Nature any such prefixed Bounds” (iii, vi,
§30, 457). For example, scholastic philosophy generally assumed that ‘rational­
ity’ defined the boundary of the species ‘man.’ Yet theological debates about
whether a deformed foetus or a changeling deserved baptism indicated that
physical appearance, not rationality, truly governed our use of this general
word (iii, vi, §26, 453–4). Analogously, a rational horse, like Jonathan Swift’s
Houyhnhmn, would never be called a ‘man’ in ordinary language (iii, vi,
§29, 456). As these examples suggest, common language contradicts the philo­
sophical belief that species are bound by metaphysical essences or even stable
definitions. The boundaries of species would seem to be dictated by language
rather than by ‘nature.’ These observations in the Essay seem to point to a
“radical arbitrariness” in not just words but concepts (Formigari 1988: 88).
Though indeed radical, Locke’s thought should not be confused with what
Gottfried Wilhelm Leibniz called the ‘super-nominalism’ of Thomas Hobbes,
who allegedly reduced philosophical propositions to mere words (Leibniz
1956: 199). Locke had set out specifically to distinguish language from truth,
and to correct the ‘abuses’ of philosophical language, not to prove that lan­
guage itself constituted how we understand the world. In some cases, Locke
acknowledged a central place for language in thought without much ambi­
guity. “Mixed modes” was Locke’s term for the complex ideas designated by
moral and legal concepts such as ‘parricide,’ ‘courage,’ or ‘incest.’ In Locke’s
view, these complex ideas were entirely voluntary constructions without any
necessary archetype in existence, as evidenced by languages that contain a
word for killing a father but no word for killing a son. In the case of such
complex ideas, the word does not just designate the complex idea but serves
as the “Knot” that holds the ideas together (1975: iii, v, §10, 434). Complex
ideas of substances, however, differ from “mixed modes” because they sup­
posedly derive from real objects in nature, and therefore do not need a verbal
“Knot.” Instead, the mind assumes that simple ideas consistently combined in

263

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

a complex idea like ‘man’ or ‘gold’ are joined by some “substratum” or “Sub-
stance” in the object itself (ii, iv, §18, 95). Not only scholastic philosophers,
that is, but also common speakers assume that their words refer to some
‘essence’ or occult property in the object itself. Although Locke inferred that
simple ideas were caused by inherent ‘powers’ generated by the atomic con­
stitution of different objects, he denied that we could know anything about
them. Given the limitations of the senses, the ‘essence’ named by words was
an inference, or what he called a “nominal Essence” (iii, vi, §2, 439).
Hence, although Locke denied the existence of metaphysical entities such
as ‘substance’ and ‘essence,’ he continued to assume that the understanding
required these concepts in order to name and sort genera and species. Moreo­
ver, the need for ‘nominal essences’ seems inconsistent with Locke’s doctrine
that words properly name only ‘ideas.’ As later philosophers complained,
Locke offered no definition of an “idea,” which he sometimes seems to equate
with a visual image in the mind and sometimes with any mental concept
(ii, i, §8, 47). He is nonetheless consistent in assuming that ‘substance’ and
‘essence’ are either indistinct ideas or not ideas at all. If the word gold signifies
the ‘nominal essence’ of a heavy yellow substance, it cannot denote a true
‘idea’ even in Locke’s ambiguous sense. In this and other ways, Locke’s con­
sideration of actual language use repeatedly led him to qualify or contradict
his assertion that words signify only ideas. Early in the Essay, Locke argued
that speakers commonly assume that words bear a “double conformity,” first
to the ideas of others and second to the reality of things in the world. This
double conformity was “a perverting of the use of Words” (iii, ii, §5, 406–7),
which properly refer only to ideas in the mind of the speaker. Later on, in his
chapter on “The Imperfection of Words,” however, he indicates that words
are properly used for “the communication of our Thoughts to others” and
“the recording of our own Thoughts” (iii, ix, §1, 476). This definition of the
function of language suggests both that we properly refer our words to the
thoughts of others and that words have an internal function of organizing our
own ideas, a crucial mental function of language that Locke had previously
only vaguely implied. Still later in the Essay, Locke seems entirely to contra­
dict his earlier statement that it is a perversion of language to assume that
words for substances must stand for “things.” “Our Names of Substances,”
he writes, are “not put barely for our Ideas,” but are “made use of ultimately
to represent Things” (iii, xi, §24, 520). According to this definition, we cannot
assume that general ideas of substances are merely “the Workmanship of the
Understanding.” We must instead make words conform as closely as possible
to the real nature of non-mental objects.

264

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

These various statements have been grist to the mill of modern scholar­
ship on Locke’s philosophy of language, which has dwelled on the question
of whether he believed that words signified ideas or things (see Kretzmann
1977; Ashworth 1981, 1989; Ott 2004; Losonsky 2007). It is perhaps best to
conclude that he genuinely contradicted himself in ways that reflect his own
divided objectives. On the one hand, as we have seen, he set out to disentan­
gle the confusion between words and things that he found characteristic of
scholastic philosophy. On the other hand, Locke was seriously committed to
the aims of modern science. This tension between his rebuttal of scholastic
philosophy and his belief in modern science led him to distinguish between
two levels of language, “Civil” and “Philosophical” (1975: iii, ix, §32, 476). The
first category, which refers to how people normally name things, accom­
plishes the main end of words, to communicate ideas to others. Though he
often sounds contemptuous of the “imperfections” sanctioned by common
language, Locke often refutes philosophical uses of language by pointing
to how words are commonly used. As noted above, Locke debunked the
definition of ‘man’ as ‘a rational animal’ by pointing to how ‘man’ is in fact
an ambiguous and contested species in ordinary speech. Similarly, the way
common speakers use ‘I’ or ‘self’ is more reliable than philosophical defini­
tions which equate personal identity with some immaterial entity apart from
consciousness and memory. In fact, “myself” is the expression used in ordi­
nary speech to refer to conscious experience and what I remember about
my past life (see 1975: ii, xxvii). Although Locke did not deny the existence
of an immaterial ‘spirit,’ this term denoted (like ‘nominal essence’) some
unknown substratum of my thoughts. Nevertheless, Locke also claimed that
philosophers should use words more accurately than common speakers do.
Philosophers were under the obligation to use words as accurately as possi­
ble to denominate ‘things’ in the world, even if this language both departed
from common usage and contradicted his insistence elsewhere that words
denominate only ‘ideas.’
In Nouveaux essais sur l’entendement humain (New Essays on Human Under-
standing, Leibniz [1765]1981), therefore, we reach a crossroads dividing sci­
ence, linguistics, and philosophy. The new science that Locke admired
would lead eventually to the taxonomies of Carl Linnaeus and G.-L. Leclerc,
Comte de Buffon, and to the new chemistry of Joseph Priestley and Antoine
Lavoisier. This emerging tradition of science would classify the natural
world and its material substratum without showing much interest in the
philosophy of language. In empiricist philosophy, on the contrary, Locke
had made language central to a whole new set of problems. He had inspired

265

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

the conviction that language not only reflected the world or the mind, but
that it ‘constituted’ reality in ways far more profound than he himself had
acknowledged.

Initial Reactions to Locke’s Essay

Locke’s argument that general words do not correspond to real species and
genera in nature, but are created instead to facilitate language, met with pre­
dictable and largely conventional hostility from philosophical and theological
conservatives (Sergeant 1697; Stillingfleet 1697). His arguments were subjected
to a more sophisticated critique by Leibniz, whose detailed reading of Locke’s
Essay in New Essays, was not published until 1765. Trained in a German tradi­
tion, Leibniz did not entirely dismiss speculations about the natural affinity
between words and nature that Locke had brushed off quickly (see Aarsleff
1982: 58–60). On the other hand, Leibniz called Boehme’s Natursprache no
less “a multitude of follies” than magic and cabbalism. He instead projected a
new philosophical language or “general characteristic” based on the artificial
characters of algebra rather than Adamic resemblances (1956: 340–2). Leib­
niz’s critique of Locke’s Essay expressed his conviction that every coherent
definition of a general concept reflected a metaphysical reality fully compre­
hended only by the all-knowing intellect of God. ‘Nominal’ definitions of man
or gold varied because this reality was viewed from the particular and limited
perspectives of human reason. Whatever Leibniz’s gestures towards some
original language of nature, he conceived of language as external to the real­
ity it attempted to define, and as capable of philosophical improvement.
Whereas Leibniz’s critique of Locke would not be widely known until late in
the century, George Berkeley’s The Principles of Human Understanding ([1710]1937)
exercised enormous influence on the immediate direction of empiricism, lead­
ing to the skepticism of David Hume. Berkeley also made strikingly original
remarks on language, though they do not lead to the radical integration of lan­
guage and thought that one might initially anticipate. Berkeley opened his trea­
tise by focusing on Locke’s use of the term “general idea.” Unable to find in his
mind any general idea of ‘man,’ for example, Berkeley concluded that the word
man could only refer to some individual mental image converted into a mental
sign for all men indiscriminately. In a Treatise of Human Nature, Hume cred­
ited Berkeley with the insight that particular ideas are made general through
the addition of “a certain term, which gives them a more extensive significa­
tion” ([1739–40]1978: 17). But this is not what Berkeley precisely says. Rather, he
argues that either an idea or a word can become general by being made the sign

266

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

of an unlimited number of individuals. For this reason, as Berkeley indicated in

the introduction to the Principles, a word can substitute for an idea, signifying
a range of individuals without evoking any particular image in the mind. As
Edmund Burke would elaborate in his Philosophical Enquiry into the Origin of Our
Ideas of the Sublime and Beautiful, the sentence I shall go to Italy next summer can
be perfectly meaningful without the speaker visualizing images of continental
travel or a warm day in Rome ([1757]1812: 327). Berkeley had evidently broken
the link between ideas and words that Locke had believed necessary to mean­
ing. Indeed, Berkeley thought of language as having many functions beyond
the communication of ideas. “There are other ends,” he wrote, “as the raising
of some passion, the exciting to, or deterring from an action, the putting the
mind in some particular disposition” ([1710]1937: 22). Locke, by contrast, had
characterized the use of language to raise emotions as an “abuse” of language
typical of “Rhetorick” (1975: iii, x, §34, 508).
Berkeley both broadened the understanding of how words are used and
potentially made language integral to thought (Formigari 1988: 130). His dis­
cussion of words during the remainder of the Principles is nonetheless anticli­
mactic, for he reverted to conventional strictures on the ‘abuse’ of language.
As language deceived us into believing such fallacies as the real existence of
material objects, he beseeched the reader to attend to his own wordless men­
tal experience rather than the words on the page. Berkeley’s insistence that
the so-called general idea was in fact a particular image made into a sign of
similar individuals reformulated Ockham’s medieval nominalism, narrowing
Locke’s much broader (if imprecise) description of the ‘idea’ as much more
than a visual image. In the end, Berkeley had not greatly extended Locke’s
ambivalent insights into the mental use of words. Language remained the­
oretically external to the understanding until the work of Etienne Bonnot,
Abbé de Condillac.

Condillac and the Constitutive Role

of Language
After Locke’s Essay, Condillac’s Essai sur l’origine des connoissances humaines
(Essay on the Origin of Human Understanding) ([1746]1971) is the most influen­
tial work in the philosophy of language during the eighteenth century (see
Chapter 9). As I will suggest, Condillac’s attempt to correct the deficiencies
in Locke’s treatment of language generated yet another set of problems and
objections. Nevertheless, Condillac gave a fresh direction to linguistic philos­
ophy in two important ways. First, he greatly deepened the role of language

267

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

in the development of understanding, believing that Locke failed to appre­

ciate the full significance of words because he had come to this subject only
belatedly. Second, Condillac pushed back his analysis of both language and
understanding to the beginnings of human society. Locke had begun with the
child’s first assembly of simple ideas into complex ideas. Condillac believed
on the contrary that a legitimate history of understanding must begin with
the very first speakers, for Locke’s child would quickly absorb assumptions
embedded in the language around him. Condillac’s historical approach to lan­
guage was not itself original. His conjectures on the origin of language drew
in part from a classical tradition dating back to Lucretius and Diodorus Sicu­
lus. Probably without his knowledge, he was also anticipated by the Anglo-
Dutch writer Bernard Mandeville, whose cynical satire on human greed and
vanity, The Fable of the Bees ([1724]1924), traced language back to the lewd ges­
tures and aggressive shouts of our primitive ancestors (ii, 286–302). Another
important and professed influence on Condillac was Bishop William War­
burton’s section on the history of written language in the second volume
of The Divine Legation of Moses (1738), which was separately translated into
French in 1744. Warburton described the evolution of pictographs to hier­
oglyphs, which he wrongly construed as ideograms. Egyptian scribes con­
trived increasingly metaphorical and enigmatic symbols for abstract ideas,
leading finally to the invention of arbitrary marks for consonants and verbs
(see Hudson 1994: 59–71). This thesis helped to inspire Condillac’s thesis that
verbal language, as well, might have begun with natural signs that became
arbitrary and abstract through a gradual process. Condillac nonetheless went
further than Warburton in arguing that language did not just signify ideas,
but was fundamental to the evolution of reason itself. “The use of signs,” he
wrote, “is the principle which unfolds all our ideas as they lye in the bud”
([1746]1971: 11).
This conviction stemmed from Condillac’s argument that arbitrary words
are no less necessary to bind together complex ideas of substance than what
Locke called “mixed modes.” Locke had called words such as parricide and
courage the “Knot” tying together an entirely voluntary assemblage of sim­
ple ideas. Was the complex idea that we call ‘fruit’ or ‘lion’ really any differ­
ent? Condillac argued that any initial perception of an object could not be
held in the memory or revived at will without an arbitrary sign. Without
conventional language, the person might recall a perception accidently by
having a sensation associated with it through previous experience. Hunger
might recall the fruit; fear might recall the lion. But these fleeting associa­
tions would never convert a perception into an idea, a term restricted by

268

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

Condillac to a perception that had become the subject of conscious reflec­

tion. Only the arbitrary sign enabled reflection by allowing the individual to
summon a complex idea as a unit rather than as a sequence of individual sen­
sations ([1746]1971: 118–19). Without this capacity, a person without language
would be unable to have clear memories or to form abstract ideas. Condillac
believed that the first perceptions would be of individual objects that could
not be made general until they were analyzed or “decompounded” into their
component simple ideas. These simple ideas could then be reassembled to
form complex general ideas. Hence, the simple idea, rather than the complex
idea, was the original product of abstraction, a procedure impossible without
arbitrary signs.
The problem faced by Condillac was to explain how humans without lan­
guage, memory, or even proper ideas could ever have invented these arbi­
trary signs. Setting aside the biblical account in Genesis, Condillac imagined
two children wandering in a wilderness without any knowledge of language.
The sensation of hunger might “put these children in mind of a tree loaded
with fruit, which they had seen the day before.” But the children would be
unable to signify their wants or how to relieve them except by “the cries
of each passion” accompanied by appropriate gestures ([1746]1971: 171–2). In
time, however, the children would associate certain cries and gestures with
their needs and the associated objects, and would repeat originally natural
signs even in the absence of these sensations. This was the birth of the arbi­
trary sign, giving human beings the power to reflect on their perceptions
and to revive them at will. Comparably with Warburton’s thesis on the evo­
lution of writing, Condillac argued that the evolution from the natural to
the entirely arbitrary sign took place over a long time. Language remained
linked for many centuries with natural cries and gestures, as evidenced by the
musical language and dancing of early Greek drama. Eventually, however,
language became both more arbitrary and more precise, losing its original
dependence on gesture, and approaching the emotionally cold yet rational
precision of modern languages, especially French.
Although Condillac had made language integral to the understanding in
a way that Locke had not, his philosophy had not broken entirely with the
prevailing conceptions of the previous era. He still criticized the abuse of lan­
guage by scholastic philosophy, urging that “We ought never to make use
of signs, but in order to express ideas which we really have in our minds”
([1746]1971: 262). Such statements are confusing in the context of an argument
that evidently sought to eliminate precisely this distinction between signs and
ideas. Condillac continued to suggest that as language grew more precise and

269

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

analytical it enabled the “progress of the mind” towards a higher degree of

“perfection” ([1746]1971: 290). Yet if it is true, as he wrote later in his Logique,
that “abstract and general terms are nothing but names” (1982: i, 388), it is dif­
ficult to determine what Condillac could have meant by “perfection” besides
the internal coherence and order of language itself, which he sometimes
compared with algebra. At times, indeed, Condillac seems to recreate Berke­
ley’s idealism in the form of a purely linguistic construction (see [1746]1971:
312). This tendency to equate “truth” with the sheer order and precision of
an inherently linguistic construction was indeed a concern for later writers
who wished to restore the link between language and an objective non-
mental reality. Nevertheless, the more immediate objection was that Condil­
lac had not found the origin of language at all, but had pushed the relation­
ship between language and reason towards a paradoxical dead-end.

Problems of Origin: Language, Mind,

and Society
Although it is legitimate to speak of a “tradition of Condillac” (Aarsleff 1982:
146–209), we should understand that his influence, like Locke’s, was extended
through a series of critiques and outright rejections. Along with Warburton,
Condillac had popularized a methodology in linguistic philosophy that the
early nineteenth-century Scottish philosopher Dugald Stewart dubbed “Theo-
retical or Conjectural History” (1811–12: v, 450). This was a technique that tried
to imagine how primitive speakers or writers could have invented such a
sophisticated and complex practice as language. Particularly after Condillac,
the emphasis of philosophy shifted dramatically from the abuse of language
to the unexpected sophistication of its very achievement, an important step
towards the creation of a specifically linguistic science. By the early nineteenth
century, methodologies that relied on speculation were being overtaken by
a new linguistic science that insisted instead on empirical fact. Ironically,
empiricism spawned its own rebuttal. But this was a later development, for
the first wave of criticism against Condillac concerned not his neglect of fact
but rather the paradoxes raised by his attempt to trace language back to its
theoretical origins.
These were paradoxes formulated most influentially by Jean-Jacques Rous­
seau in Discours sur l’origine et les fondements de l’inégalité parmi les hommes, now
known in English as A Discourse on Inequality ([1755]1985; see Chapter 9). Rous­
seau was personally acquainted with Condillac and had been instrumental in
the publication of his Essai. In the Discours, however, Rousseau alleged that

270

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

Condillac had failed to address serious and perhaps insurmountable problems

in his history of language. First, “he assumes what I call into question, namely
some sort of society already established among the inventors of language”
([1755]1985: 92). Rousseau insisted on the contrary that human beings lived
largely in isolation from each other in the state of nature and would have
lacked the language needed to found societies in the first place. As Rous­
seau asked “Which was more necessary, a society already established for the
invention of language, or language already invented for the establishment
of society?” (pp. 96–7). A second and “even worse” problem stemmed from
Condillac’s own argument that reason was impossible without language, a
position that Rousseau evidently regarded as proved. “If men needed speech
in order to learn to think,” Rousseau objected, “they needed still more to
know how to think in order to discover the art of speech” (p. 93). As Rousseau
portrayed people in the state of nature as little more than animals directed
by present sensations and instincts, it struck him as inconceivable that they
could have invented general and abstract ideas. “General ideas can only be
introduced into the mind with the assistance of words; and the understanding
can grasp them only by means of propositions” (p. 95).
Rousseau’s critique of Condillac does not seem entirely fair. Condillac had
indeed recognized that the interdependence of language and reason created a
paradox. He attempted to resolve this difficulty by arguing that the first arbi­
trary words could have arisen from natural cries and gestures, sparking a long
historical process leading to general words and a grammatical language. But
Rousseau was correct in detecting that the methodology popularized by Con­
dillac persistently raised problems of origin, constantly leading philosophers
to assume some prior language or reason before their supposed invention.
Ironically, Rousseau’s own conjectural history of speech, Essai sur l’origine
des langues (Essay on the Origin of Languages), first published posthumously
(1781a), is plagued with precisely this kind of problem, for his attempts to
locate some moment just before language and reason always seem to imply
that they already existed (Derrida 1967/1974: 165–268). Rousseau is notoriously
inconsistent, but even such a highly exact philosopher as Adam Smith fell
into similar contradictions in Considerations on the First Formation of Languages
([1759]1983), the work which Stewart had in mind when he coined the term
“conjectural history” (1811–12: v, 450). Smith began this short tract by denying
that the problems raised by Rousseau about the original formation of species
and genera were really so difficult to resolve. The first speakers would have
merely extended their name for a particular cave or tree to all similar objects.
Thereafter they would have distinguished between, for example, green trees

271

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

and withered trees by inventing an adjective. It nonetheless occurred to Smith

that the creation of a separate word to describe a quality of an object would
require “a considerable degree of abstraction and generalization” ([1759]1983:
206) impossible among the primitive inventors of language. These first speak­
ers would have overcome this difficulty, he assumed, by indicating various
qualities and relations through varying the root word in some euphonic and
thus easily remembered way (like canus to cani), the origin of inflected lan­
guages such as Greek and Latin. It is unclear, however, why the invention of
an inflection requires less abstraction than the invention of a separate word for
an adjective or preposition (Land 1977: 684). Smith went on to weigh the pos­
sibility that in fact verbs would have been invented before nouns, as primitive
speakers would have been most attentive to active events such as the weather
and threatening animals. This second thesis on the origin of language led to
his similarly problematic conjecture that the intellectual effort of abstracting
the subject from the verb would have been avoided by adding endings to the
verb stem. The product of this evolution, Smith concluded, would have been
synthetic languages characterized by an increasingly complicated apparatus
of cases and conjugations, Greek being a notorious example. Synthetic lan­
guages must have preceded analytic languages. With the mixture of peoples
through migration and conquest, however, the grammatical particles of mod­
ern analytic languages would have been invented for the sake of “simplifica­
tion” ([1759]1983: 220–1). If separate adjectives, prepositions, and pronouns are
simpler than variations on a root word, we might ask, then why were they
not invented in the first place? In short, Smith is led from a straightforward
and confident beginning into increasing perplexity and self-contradiction.
The theoretical difficulties exhibited by Smith’s Considerations led to a
revival of a thesis that we do not usually associate with the Enlightenment,
the divine origin of language. Rousseau’s suggestion in the Discourse that he
could not conceive how language could have been invented by human means
([1755]1985: 96) was endorsed in the article “Langue” by Nicolas Beauzée and
Jacques-Philippe-Augustin Douchet (1765) in that flagship of Enlightenment
thought, the Encyclopédie. The divine origin of language was later main­
tained in respected discussions by Johann Peter Süßmilch (1766), James Beat-
tie (1788b), and Hugh Blair ([1783]1965). Another version of this rejection of
Condillac’s thesis on the natural origin of language was presented by James
Burnett, Lord Monboddo’s Of the Origin and Progress of Language (1773–92).
Stressing the insurmountable difficulty of imagining how language could
have evolved from natural cries or gestures, Monboddo set out his lengthy
case that language must have been invented not by God, but by ingenious

272

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

Egyptian philosophers. Monboddo found little difficulty in resolving Rous­

seau’s first question about the precedence of language and society: notori­
ously, he urged that even orangutans had “societies” without language,
and could even develop human-like societies with the invention of speech.
Nonetheless, Rousseau’s second question concerning the precedence of lan­
guage and reason struck Monboddo as irresolvable without metaphysical
insights well beyond the capacities of either orangutans or primitive humans.
Unlike his fellow Scot, Adam Smith, Monboddo drew evidence not just
from the learned languages, Greek and Latin, but from recent grammars of
North American languages such as Huron. His reading of these grammars
convinced him, in contrast with Smith, that primitive humans could not
graduate beyond the limitless accumulation of words signifying innumerable
perceptions and even entire propositions. ‘Great bear’ and ‘small bear’ were
allotted separate expressions. ‘He is handsome’ and ‘he is not handsome’
were signified by utterly different words (1774: i, 351–2, 365). Only a school of
philosophers could have decomposed this bewildering mass of vocabulary
into a system of grammatical parts, abstracting adjectives from nouns, sub­
jects from verbs, and so forth.
The fundamental problem for Monboddo, as for all empiricist philosophers
since Locke, turned on the problem of how words or mental signs for single
perceptions could have crossed the barrier from raw sensory experience to the
abstractions and generalizations fundamental to rational discourse. Debunk­
ing the scholastic language of genera and species, of essences and substance,
had not solved this problem, but had instead driven philosophers back into
an evidently endless search for origins. Conjectural history also raised the
possibility that, again, seems inconsistent with many modern assumptions
about the Enlightenment: since reason was entirely dependent on language,
every language created a unique understanding of the world that had little to
do with some objective or universal ‘truth.’

Relativism and “the Natural Order of Ideas”

Having denied any innate capacity in the mind to organize sensations into
species and genera, empiricist philosophers quickly seized on the problem
of how primitive humans could even have begun the process of analyzing
experience into a grammatically organized language. Interest was intense
in William Cheselden’s widely reported removal of cataracts in 1729 from a
13-year-old boy who testified that he first saw only a shifting kaleidoscope of
colors (Voltaire [1738]1992: 319–20; Condillac [1746]1971: 184–5; Reid 1863: 136).

273

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

This testimony seemed to confirm the belief that the joint evolution of rea­
son and language must have followed the decomposition of this manifold to
the order reflected in modern language. But where had this decomposition
begun? Answers to this question ranged from an extreme relativism to the
reaffirmed belief that the human mind possessed an innate capacity to reduce
sensations to the grammatical order of modern speech.
The case for extreme relativism was presented in Pierre-Louis Moreau
de Maupertuis’s Réflexions philosophiques sur l’origine des langues (Philosophical
Reflections on the Origin of Languages) ([1748]1768). Though himself a natural
scientist, Maupertuis indicated that even modern science derived from the
evidently arbitrary decision of the primitive inventors of language to decom­
pose the world in one way rather than another. In the language of western
science, for example, tree refers to an object of a certain shape and size, not to
contingencies such as the green, brown, or orange color of these objects. A
different initial decision, that for example all trees must be green, would have
led to an entirely different categorization of the world (i, 272–3). All of sci­
ence has been determined by these decisions. Hostility to Maupertuis’s thesis
in the French linguistic tradition is hardly surprising (Turgot [1805]1971; see
Formigari 1993: 36–40; Ricken [1984]1994: 111–33). The French tradition gen­
erally followed Condillac in holding that the French language, above all, had
imposed an exemplary rational order on the manifold of experience. While
this order did not necessarily map out any outward reality, the rationality of
this order would not be haphazard in what Condillac called a “well-formed”
language (1982: 400).
According to Gabriel Girard in Les vrais principes de la langue françoise (The
True Principles of the French Language) ([1747]1982), the superior rationality of
French derived from its particular origin. Girard claimed to be the first writer
to have divided languages into distinct types (p. 23). French exemplified the
class of “analogic” languages which follow “l’ordre naturel … des idées”
(p. 24). He denied that French could have had the same origin as Latin, which
was a “transpositive” language shaped by the impulses of the imagination
rather than the order of nature. While the mixture of tongues had produced
what Girard called “amphilogical” languages, the presence of the definite arti­
cle in analogic languages, along with the lack of the cases typical of transpos­
itive languages, indicated the quite different procedures of those who first
named the world. The first speakers of analogic languages initially noticed
the being of things in general, as signified by the definite article. French orig­
inated with some word like le (‘the’), a word for ‘being’ later attached to spe­
cific objects, whereas the creators of transpositive languages began by naming

274

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

these objects alone. Girard’s insistence that the recognition of ‘being’ in gen­
eral was the first step towards any truly rational language, is only one indi­
cation of an Aristotelian conservatism that also led him to deny that words
themselves played any constitutive role in the creation of rational thought.
Words mirrored, rather than influenced, different modes of understanding.
The free-thinking philosophe Denis Diderot was, not surprisingly, more
sympathetic to the more radical implications of Condillac’s thesis. Diderot’s
Lettre sur les sourds et muets (Letter on the Deaf and the Mute) ([1751]2010) is a
remarkably complex and suggestive discussion of the formative impact of
language on thought, though he too wished to maintain that French recorded
the ‘natural order of ideas.’ Diderot proposed that there are different ways of
understanding this natural order. If we think of the order in which languages
must have been created, then adjectives must come before nouns, for nouns
(which are all general) could only have been named after the decomposition
and abstraction of the qualities that constitute individual objects. Alterna­
tively, we might consider how deaf and mute people communicate, though
it would be difficult in these cases to distinguish between what is ‘natural’
and what they have learned from surrounding conventions. While both these
methods are revealing, it is best to begin with the insight that all languages
impose a particular structure on thought, which Diderot compared to a well
or badly organized library of sensations. Like no other language, French has
reduced this library to an order that, like the great Encyclopédie he edited, ren­
dered thinking orderly and systematic ([1751]2010: 222–4).
For all their differences, Girard and Diderot display common characteris­
tics of French linguistic thought in the wake of Locke and Condillac. First,
both think of the sentence, not the word, as the significant unit of language,
insisting that every grammatical unit of a sentence is meaningful only in the
context of a proposition that may be either well or badly organized. Second,
although Girard and Diderot were less inclined than Rousseau to lament the
cold rationality of French civilization, they did agree with their iconoclastic
contemporary that the precision of French had come at the expense of the
passionate capabilities of less rational tongues. Above all, Girard and Diderot
were thorough empiricists who denied that language and thought were
influenced by anything but sensory information. As we have considered, this
empiricist model of mind exemplified by Locke increasingly led philosophers
of language into the paradoxes identified in Rousseau’s Discourse ([1755]1985).
Short of reviving a theory of divine origins, or postulating like Monboddo
that language was invented by a philosophical academy, later writers saw the
need to reassert some innate capacity of the human mind to form language.

275

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

Conservative Reactions against Strict

Empiricism
While this conservative reaction could not return completely to a scholastic
faith in the homology between language and metaphysical reality, it could
insist that the Aristotelian tradition had been correct to place its faith in the
essential rationality of language. As James Harris wrote in Hermes: or, a Phil-
osophical Inquiry concerning Language and Universal Grammar, “the ORDER of
the Termes, as they stand ranged by the old Grammarians, is not a fortuitous
Order, but is concomitant to our Perceptions” (1751: 138). The inherent ration­
ality of all human language is proved by the capacity of the same speaker
to learn different languages. Indeed, even communication between speak­
ers of the same language would be impossible without a range of basic and
unlearned assumptions. “Is it not marvelous,” Harris argued, “there should be
so exact an Identity of our Ideas, if they were only generated by sensible Objects,
infinite in number, ever changing, distant in Time, distant in Place, and no
one Particular the same with the other?” (pp. 398–9). Harris denied that any
language could “express the Properties and real Essences of things” (p. 335), an
indication of Locke’s impact. But he insisted also that it would be impossible
to organize or remember the diversity of sensory objects without the inher­
ent power of the mind to impose “Internal Form” on exterior matter (p. 375).
This was a considerable step beyond strict empiricism, and even recalls the
innate mental powers of Aristotelian philosophy. Yet Harris was admired by
a great founder of comparative linguistics, Sir William Jones, who called him
“the most sagacious philosopher of the age” (1771: xv).
The aim of refuting the skeptical implications of empiricism also motivated
the Scottish philosopher Thomas Reid, a major figure in the “common sense”
school of philosophy that emerged to counter Hume’s threat to both philo­
sophical and Christian orthodoxy. Like Harris, Reid argued in Essays on the
Intellectual Powers of Man that the universal translatability of languages, along
with the very fact of communication among different individuals, proved that
there were “common principles of human nature” that made language pos­
sible ([1785]1971: 460). What empiricists such as Locke, Berkeley, and Hume
regarded as errors in common language in fact represents the inherent prin­
ciples of “common sense” that neither language nor the understanding can
do without. A child or the most ignorant person naturally and immediately
believes that his words conform with the usage of others, and that they signify
real objects in the world. So far from being a “perversion” of language, Locke’s
“double conformity” belonged to a universal “common sense” that empiricist

276

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

philosophers denied at the expense of making their own language incompre­

hensible. Hume’s denial of cause and effect, for example, was itself an abuse
of language. Reid did not pretend to understand how even children or savages
achieve “abstract general notions,” without which “there can be neither rea­
soning nor language.” Nonetheless, Locke and later empiricists had failed to
explain the existence of these notions, which were utterly inconsistent with
their insistence on the primacy of “simple ideas” ([1785]1971: 475–82).
In the words of Johann Gottfried Herder in his Abhandlung über den
Ursprung der Sprache (Essay on the Origin of Language), empiricist accounts had
led the history of language into an endless swing around a “perpetual merry-
go-round” ([1772]1986: 120). One could escape this merry-go-round by two
means. One could return to a divine origins theory of language, as argued
by Herder’s compatriot Süssmilch. But this thesis itself implied, as Herder
insisted, that Adam was already rational enough to understand what God
was teaching him. Alternatively, one could infer that humans possessed spe­
cial and inherent powers that made language possible. Herder combined a
version of this second alternative with a methodology that, in contrast with
Harris, remained indebted to the methodology of conjectural history. To this
extent Herder remained in the tradition of Condillac, though he dismissed the
thesis that language had evolved from natural cries and gestures as “childish”
and “hollow” (pp. 110–11). Like Rousseau, he claimed that Condillac had failed
to resolve the conundrum that speech and reason were mutually dependent.
Equally, however, Herder disdained Rousseau’s alleged reduction of human
nature to the level of brutes, an error epitomized by Monboddo’s argument
that even orangutans could be educated to speak and reason. Humans pos­
sessed a power to create language as natural to us as the instinctive and inar­
ticulate cries of animals. Herder resisted the conventional solution of calling
this special power ‘reason,’ a term that was itself an ‘abstraction.’ What “rea­
son” actually referred to was “the total arrangement of all human forces, the
total economy of his sensuous and cognitive, of his cognitive and volitional
forces” (pp. 109–10). Insofar as there was any term for the special power of
the human mind, it was “reflection,” which meant the capacity of the human
mind to identify and name some rudimentary object in the sensory manifold,
or what Herder called “the vast ocean of sensations” (p. 115). Here Herder was
closer to Condillac than he admitted, for the French philosopher had similarly
distinguished between a mere ‘perception’ and an ‘idea,’ the latter being a
sensation that was held in the mind through the power of reflection.
The necessity of particular attention to some part of the sensory manifold
had been previously discussed by David Hartley in a work of unstinted and

277

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

mechanistic empiricism, Observations on Man ([1749]1966). Hartley wondered

how the child could distinguish “essentials” from “non-essentials” in isolating the
object named “Nurse” which appeared sometimes in one dress and sometimes
in another, sometimes in one room and sometimes in another. Hartley specu­
lated that the milk of “Nurse” would draw the child’s attention to her particular
identity, dividing her from an otherwise bewildering sensory context (pp. 271–2).
Returning to the first language of the primitive human being, Herder ([1772]1986)
argued instead that only “sound” could serve as the “distinguishing mark” that
divided the individual from the “ocean of sensations.” His example was the bleat­
ing of a sheep: “The sound of bleating perceived by a human soul as the distin­
guishing mark of the sheep became, by virtue of this reflection, the name of the
sheep, even if his tongue had never tried to stammer it” (p. 117). Although Herder
implied that the first word for ‘sheep’ sounded something like ‘baa,’ this was not
his main point. His point was rather that the sound of bleating served as a mental
‘word’ that, while private in origin, prepares the way for public language.
Herder reflects an important trend in late eighteenth-century philosophy in
elevating sound over vision as the fundamental sense in human understand­
ing. Contrary to what has been often assumed, the ascendency of literacy
and print culture led to increased appreciation of oral over visual language
(Hudson 1994: 162–6). A previous tendency to conceive ideas as visual sensa­
tions had led Berkeley, in particular, to deny the existence of ‘general ideas’
because he could mentally visualize ‘man’ only as a particular individual.
Herder ([1749]1966), on the contrary, argued that a mental image could never
serve as the basis of language. Only a sound could open “the gateway to [the]
soul,” establishing an essential bridge between the self and the world. Vision
is too “cold and aloof” to serve as this gateway, while touch “overwhelms”
by being too connected with the self. All language, whether mental or social,
must be based on sound, “the middle sense” (pp. 142–4). As Herder realized,
this hypothesis led to the problem of how humans came to create words for
visual or inaudible perceptions. It was here that his description of “reason”
as “the total economy of [man’s] sensuous and cognitive … forces” became
important. Locke had famously satirized a blind man’s analogy between the
color scarlet and “Sound of a Trumpet” (1975: iii, iv, §11, 425). For Herder, on
the contrary, this analogy was not only entirely human but also the basis for
the extension of originally audible signs to all kinds of non-audible percep­
tions. The initially audible mark of the sheep’s bleating, to use his example,
would be naturally transferred through the “sensorium commune,” the inter­
connected network of the human senses, to the sheep’s whiteness and soft­
ness in both the primitive mind and the first languages ([1772]1986: 139).

278

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Locke and Reactions to Locke, 1700–1780

An attractive virtue of Herder’s thesis was that it mediated between posi­

tions that had previously seemed difficult or impossible to reconcile. While
Locke and strict empiricists like Berkeley and Hartley had regarded language
as distinctly private, for example, Condillac had insisted that language could
only have emerged through a history of dialogue. Language was fundamen­
tally communal, a belief that inspired Condillac’s depiction of each language
as “the picture of the character and genius of every nation” ([1746]1971: 299).
According to Herder, the seeds of language were indeed private – the iden­
tification of the “distinguishing mark” – but this characteristic act of human
reflection made possible a communal dialogue governed by the particular
environments and needs of different peoples. The differing languages of
Arabs, Caribs, and Hurons, to cite some of his examples, reflect the “custom,
character, and origin of the people” ([1749]1966: 155). This argument might
seem to imply that each language created quite a distinct version of reality,
as suggested by Maupertuis or, in a rather weakened version, by Herder’s
fellow German Johann David Michaelis (see Chapter 9). In an essay that won
a prize from the Berlin Academy in 1759, Michaelis argued that every lan­
guage had been formed according to the needs and experience of its speakers,
which meant that no two languages were exactly translatable and that the
etymology of every language traced a unique linguistic heritage. Michaelis
nonetheless believed that language should reflect ‘truth,’ warning against
the philosophical ‘mistakes’ embedded in etymologies (1769). Herder agreed
with Michaelis that each language embodied a unique cultural inheritance.
He was nonetheless less interested in the capacity of language to reflect some
outward physical reality than in its expression of an inward human reality. As
all languages had been created according to common principles unique to
all human beings, all represented local translations of “the universal voice of
nations” ([1772]1986: 159). The originally consistent procedure of all humans
in creating languages meant that any language could be learned, and that we
recognize in any language an innately human way of organizing and articu­
lating experience. “Invention of language,” he wrote, “is … as natural to man
as it is to him that he is a man” (p. 115).

Conclusion
Although the debate traced briefly in this chapter is diverse, complex, and con­
tradictory, all the theories that we have considered find a common source in
John Locke’s Essay concerning Human Understanding. Locke’s Essay is perhaps
best approached as a radical and richly suggestive experiment. His professed

279

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press

 Nicholas Hudson

goal of merely correcting the ‘abuse’ of language in philosophy, and par­

ticularly the misconception that words signified metaphysical ‘essences’ or
‘substances’ in the non-mental world, led ironically to a much more integral
role for words in the formation of the human understanding. The empiri­
cist tradition that followed Locke extended the role of language in the mind
well beyond what he himself had admitted or even welcomed. This tradi­
tion challenged Locke across a wide range of issues. As Condillac and others
argued, the study of language must begin not with the mind of the child, but
with the historical origins of language itself. Eighteenth-century philosophy
of language became closely linked with the history of society, and with a
wide range of social, national, and political ideologies. Similarly, later authors
argued that language cannot be understood merely as signifying private expe­
rience, but must have emerged through dialogue and meanings developed by
whole communities rather than individuals. Nor could the creation of signs
have been ‘arbitrary,’ but must have had some foundation in motivations of
need, desire, or passions of various sort (see De Brosses [1765]1801: 4–5).
Strictly empiricist accounts of the origin of language would extend into the
nineteenth century, as in the work of Destutt de Tracy and Bentham. These
philosophers continued to explore the question of how language could be
explained without assuming the existence of anything except sensory percep­
tions organized by the association of ideas and driven by basic drives which
prompt conscious reflection. Yet an important consequence of developments
in linguistic philosophy during the eighteenth century was revival of the belief
that language could not be explained without some foundational capacity apart
from mere sensory information (see Formigari 1993: 40–62; Ricken [1984]1994:
191–225). This departure from empiricist tradition should lead us to question
conventional assumptions about eighteenth-century thought after Locke. The
so-called ‘Enlightenment’ or ‘Age of Reason’ in fact challenged and undermined
confidence in the connection between words and non-mental reality, and in the
possibility of a generalized description of language or of reason apart from the
particular historical experience of each linguistic community. The thesis that
language and reason cannot be regarded as distinct activities has remained a
profound and lasting contribution of post-Lockian philosophy. Nevertheless,
the comparative, structural, and cognitive methods of modern linguistics were
made possible only by a conservative reaction against the skeptical implications
of empiricism. This reaction reaffirmed the innate and essential power of the
human mind to form language and experience in ways subject to scientific anal­
ysis. Humans did not just learn to speak through experience. Language repre­
sents an original and integral capacity of human nature.

280

https://doi.org/10.1017/9780511842788.017 Published online by Cambridge University Press