On Defining the Proper Name
U. OF F. LIBRARIES
University of Florida Press
o / / EDITORIA]
c 3 Humanitie
T. WALTER HERBERT, Chairman
Professor of English
J. WAYNE CONNER
Professor of Romance Languages
RICHARD H. HIERS
Professor of Religion
G. PAUL MOORE
Professor of Speech
CHARLES W. MORRIS
Professor Emeritus of Philosophy
Professor of Music
C. A. ROBERTSON
Professor Emeritus of English
MELVIN E. VALK
Associate Professor of Germanic
Library of Congress Cataloging in Publication Data
Algeo, John 1930-
On defining the proper name.
(University of Florida humanities monograph no. 41)
1. Names, English. 2. English language-
Semantics. 3. English language-Noun.
I. Title. II. Series: Florida. University,
Gainesville. University of Florida monograph.
Humanities, no. 41.
PE1578.A2A5 428'.1 73-9849
COPYRIGHT 1973 BY THE STATE OF FLORIDA
BOARD OF TRUSTEES OF THE INTERNAL IMPROVEMENT TRUST FUND
PRINTED BY THE STORTER PRINTING COMPANY
SOME YEARS AGO at an annual meeting of the Ameri-
can Name Society, one of the papers was followed
by discussion of uncommon acerbity. As an organization in which
professionalism has not yet replaced gentleman-scholarship, the
Name Society's meetings are usually marked by polite exchanges
springing from a determination to find interest and significance in
whatever paper is read before it. On this occasion, however, tem-
Spers strayed, if they were not quite lost, and there was a sharp ex-
-,- change of words over the question of whether or not the subject
matter of one of the papers could appropriately be called "names."
S The disagreement arose chiefly from the fact that those involved
had different, although not clearly formulated, notions of what a
name is. Because they started from different general positions,
there was little hope of their arriving at the same particular end.
Quarrels often spring from unclear or conflicting definitions, so
the groundwork of any study must be the definition of its object.
In recent years onomastics has had several theoretical isagoges that
focus on the problem of defining names, for example, those by
Farhang Zabeeh (1968), Holger S0rensen (1963), and Ernst Pul-
gram (1954); the oldest, and in many ways still the best, of these
is Sir Alan Gardiner's study, originally published in 1940 and only
slightly revised for its second edition (1954).* Although the views
expressed in the present study are closest to those of Sir Alan, I
believe he was led astray by an inadequate view of the structure of
In this study, the assumption is made that language must be
accounted for as a series of interconnected but autonomous levels
(for which the terms stratum and plane are used interchangeably
in the following pages). This work is, however, no more an ex-
*References begin on p. 89.
plicitly stratificational treatment of names than it is a systemic
treatment or a transformational one. Its conclusions are not de-
pendent on any particular theory or formalism, although it does
assume that language is most adequately accounted for as an inter-
locking set of discrete systems. From this standpoint, one of the
chief faults of most prior treatments of names is the assumption,
often unconscious, that names are a unified monostratal phenom-
enon. The main goal of this study is to correct that apparent mis-
I am grateful to a number of persons for their help in the prepara-
tion of the study: Robert H. West and James B. Colvert, past and
present chairmen of the English Department of the University of
Georgia, for support during the summer of 1972 to complete the
writing; Robert H. Longshore for help in providing clerical assist-
ance; April Maddox for her skillful translation of the manuscript
into typescript; and a number of persons who read the study and
provided valuable criticism, especially Adele S. Algeo, O. C. Dean,
James B. McMillan, and Thomas Pyles. I have benefitted greatly
from the suggestions, examples, corrections, and objections they
offered; the errors and misinterpretations that obstinately remain
are mine alone.
1. The Problem of Defining the Term Name 1
2. Orthographical and Phonological Names 14
3. Morphosyntactic Names 20
4. Referential Names 42
5. Semantic Names: The Degree of Meaning 53
6. Semantic Names: The Kind of Meaning 68
The Problem of Defining the Term Name
T E STUDY of names is coeval in the Western world
with the study of language itself. Indeed, it might
be said that European linguistics is the child of onomastics, for the
oldest extant Greek work on language is Plato's Cratylus, an inquiry
into the relationship between words and things, in particular
whether names are in some sense naturally connected to those who
bear them or are assigned by convention merely. The dialog begins
with an argument between two friends of Socrates, Cratylus and
Hermogenes, about the latter's name, in which Cratylus maintains
that Hermogenes cannot possibly be his companion's true name. He
insists there must be a natural appropriateness of the name to the
thing it designates; however, Hermogenes means 'born of Hermes,'
the god of cleverness, money, and good fortune, and is therefore
clearly inappropriate for such a dull, impecunious, and wretched
fellow. The opening argument of the Cratylus is, to be sure, a pon-
derous, academic joke that Cratylus is using to psychh out" his
companion, the more easily to get the better of him in the serious
discussion that is to follow. Behind the etymological jest about
Hermogenes' name, however, there lies the serious question of
whether names in general can be said to have meaning or whether
they are empty labels for things. The debate between the natural-
ists, who saw the name as revelatory of the thing named, and the
conventionalists, who saw the name as an arbitrary designation,
continued through the following centuries and in somewhat altered
terms is still very much alive today. In onomastics, as in much else,
it can be said that there are few new answers and no new questions.
The study of names began in the ancient world, and if since then
we have seen farther, it is, as Newton observed and Bernard of
Chartres before him, because we stand upon the shoulders of giants
ON DEFINING THE PROPER NAME
The history of the grammatical analysis of proper names, whose
beginnings can be seen in the Cratylus, is part of the long and
tangled story of the development of part of speech systems in the
European tradition (Michael 1970; Robins 1967). Within that tra-
dition, proper names seem first to have been recognized as a distinct
linguistic category by the Stoic grammarians. Earlier, the undiffer-
entiated mix of syntax, rhetoric, and logic, out of which the study
of language was to crystalize in several directions over the cen-
turies, made do with a single term, onoma, where later gram-
marians would want at least three: name, noun, and subject. Thus
Plato, in both the Cratylus (399B, 425A) and the Sophist (262A-
263D), uses onoma indifferently for name or noun or, as the desig-
nation of one of the two main parts of a sentence, for subject. Part
of the Stoic improvement in discrimination was to set up the onoma
proper name' and the prosegoria 'common noun' as major word
classes, based on the semantic contrast between "individual qual-
ity" and "common quality" (Robins 1966:12). The two sorts of
words were regrouped as subclasses of the same major part of
speech by Dionysius Thrax. In his short Art of Grammar, probably
the most influential textbook ever written, Dionysius says there are
two kinds of onoma: the onoma kyrion, which is applied to an in-
dividual, and the onoma prosegorikon, which is applied commonly
(Davidson 1874:331, 333). In this distinction, as in many others,
Dionysius set the fashion for the multitudes who followed him. The
two works that were instrumental in transmitting Dionysian gram-
mar to the Middle Ages and the Renaissance, Priscian's Institu-
tiones Grammaticae and Donatus' Ars Minor, echo the Greek
grammarian in subdividing nouns as nomen proprium and nomen
appellativum, the name respectively of one and of many. Priscian
expresses the difference thus: "Appellativum naturaliter commune
est multorum . ut 'animal,' 'corpus.'... Proprium vero naturaliter
uniuscuiusque privatam substantial et qualitatem significant et in
rebus est individuis, quas philosophy atomos vocant, ut 'Plato,' 'Soc-
rates'" (Keil 1864:2.58-59).
The classical beginnings of onomastic study provide in outline
the structure followed by both grammarians and philosophers to
the present time. Names are a kind of noun, a product of the first
subcategorization of nouns into two groups: proper and common.
Proper nouns are stated to have individual application, and com-
mon nouns general application, being the names respectively of
DEFINING THE TERM Name
one and of many. The only significant departure from Dionysian
grammar in the ancient world was that of Varro, who in the
De Lingua Latina (8.45) set up a rigorously formal part of
speech system and subdivided nouns into four classes: vocabula
'common nouns' like shield or sword, provocabula 'pro-common-
nouns' like who or what, nomina 'proper nouns' like Romulus or
Remus, and pronomina 'pro-proper-nouns' like this or that. Al-
though he is not explicit about the criteria for distinguishing his
subclasses, it is clear that Varro did not have recourse to the
individual-general or one-many contrast in his analysis. Rather
the principle that distinguishes the two nomen-classes from the
two vocabulum-classes is quite clearly the feature of definiteness, to
be discussed in chapter 3. In this, as in much of his grammatical
thought, Varro shows a refreshing originality, which was not to be
emulated by those who came after him. The Dionysian approach
has dominated onomastic study for well over two thousand years.
Not until the last hundred years has it been seriously challenged.
In these latter days, however, several new views have been ad-
vanced, but none generally accepted, so that in the present state of
onomastic study there is no consensus on how to define the object
of the discipline.
In view of the fact that names have been studied in a continuous
tradition for twenty-four hundred years or so, it is perhaps remark-
able that there is now little agreement about what a name is. There
are, of course, many definitions. That, indeed, is the problem. The
term name has been defined in diverse, often mutually incompatible
ways; as a result, some arguments about names spring from incon-
sistent or unclear notions of what is being discussed. Some notion
of the variety of things that find their way under the umbrella of
onomastics can be gathered from the more than five hundred entries
in Witkowski's (1964) glossary of German onomastic terms, most
of which can easily be matched with English equivalents.
Part of the present uncertainty about what counts as a name and
what does not is due to the long dominance of the classical tradi-
tion and its unthinking acceptance by succeeding generations of
students. As Br0ndal (1948:58) has observed, "La notion de nom a,
comme celle de nombre, sembl' si bien connue et si imm6diatement
comprehensible a la plupart des grammairiens qu'ils ne se sont pas
donnes la peine de lui consacrer de definition." But the notion of a
name, far from being obvious, is so unclear that one theorist (Utley
ON DEFINING THE PROPER NAME
1963:150) has questioned whether it would not be better in general
to deal with various kinds of names, such as place names and per-
sonal names, which he believes can be defined more easily, and to
leave the existence of a unified category of proper names an unre-
solved issue. However, as long as there is no clear sense of what
the term proper name means, onomatologists are in the position of
not knowing what they are talking about. This is, as Quine (1963:
47) has noted, not an untenable position, although certainly an
Part of the confusion about names is attributable to onomastics
itself, for the study of names has in practice embraced two rather
different objects: proper names and terminologies. A terminology,
also called a nomenclature, is a semantically coherent set of ex-
pressions used to designate the members of a class of objects, ac-
tivities, or so forth; it is the group of terms used in talking about a
particular subject; it is the words, usually common nouns, that
form a lexical set in the sense of Halliday (1961:276). Given the
uncertainty about proper names, the distinction between them and
a terminology can be illustrated more readily than defined. Refer-
ence to horses will do as well as any subject. On the one hand, in
dealing with horsey matters, there are such terminologies as are
exemplified by the terms palomino versus pinto, dray horse versus
carriage horse or saddle horse, mudder versus trotter or charger,
and stallion versus mare or gelding. On the other hand, there are
proper names like Bucephalus, Black Beauty, Man o' War, Dobbin,
Pegasus, and Seabiscuit. It is obvious that the two objects of ono-
mastic study are of different kinds, even though they are both
called "names" and sometimes overlap. The overlap can be seen in
the proper names for city streets, which may involve a terminology
by which streets are distinguished from avenues according to the
direction they run, north-south or east-west. Even when the choice
of term is not predictable from the compass, common nouns like
street, avenue, boulevard, lane, road, circle, way, path, alley, and
terrace constitute a terminology whose members may enter into
proper names. Similarly, personal names are one kind of proper
name, but they include titles like Mr., Mrs., Miss, Ms., Dr., and Rt.
Rev., which form a terminology. Personal names themselves may
be based on a terminology, like those derived from the days of the
week. Robinson Crusoe's Friday, who was so called "for the mem-
ory of the time," and Tuesday Weld, for whatever reason she was
DEFINING THE TERM Name 5
so called, come to mind, although better examples are the African
day-names that turned up among the black population of the New
World (DeCamp 1967). Originally they were personal names that
indicated the sex of the bearer and the day of the week on which
he was born: Quasheba 'Sunday,' Juba 'Monday,' Beneba 'Tuesday,'
Cubba 'Wednesday,' Abba 'Thursday,' Phibba 'Friday,' and Mimba
'Saturday,' to cite the female forms.
In general, however, the distinction between proper names and
terminologies is clear. Although proper names and the common
nouns that make up a terminology are both called names-in
ordinary language and in technical onomastic use-this study is
concerned only with the former, and therefore name can hereafter
be understood unambiguously as short for proper name. Termi-
nologies are no less legitimate subjects of onomastic study, but they
are not under study here.
There are other uses of name in particular "registers," or areas of
discourse, with which this study will not deal but whose existence
needs to be recognized to avoid possible confusion. For example, in
the law there is a special kind of name, the trademark or brand
name (Berle and de Camp 1959:450-78); Frigidaire is a trademark
name, sometimes used illegally in a generic sense, whereas refriger-
ator has never been a trademark and Kelvinator is so exclusively.
A trademark may be integrated so thoroughly into the vocabulary
as the generic term for an object that few speakers are aware of its
special legal status. Familiarity with the trademark and lack of any
convenient generic term to compete with it account for the fact
that a brand name like zipper is taken over by the general vocab-
ulary. Alternative generic terms, such as interlocking slide fastener,
are too bureaucratic in their associations. They sound like govern-
mental gobbledegook and are so treated by the ordinary man, who
knows enough to call a zipper a zipper. Similarly aspirin, although
originally a trademark, has passed completely into the public do-
main, happily replacing acetyl spiraeic (or salicylic) acid as the
In countries such as France, there are official lists of personal
names which may include Louis but not Lance, thus giving one and
not the other official status as a "name" (Ashley 1971). The motive
behind governmental control of naming practices may be either to
prevent the use of what are deemed unsuitable names or to en-
courage greater diversification of names for clarity of reference.
ON DEFINING THE PROPER NAME
Thus governments have encouraged the disuse of names like Cow-
ard, Hogg, and Krapp, which has not, however, prevented them
from being proudly borne by distinguished persons. When a French
court forbade Gerard and Paulette Trognon to bestow their sur-
name on their adopted son, on the grounds that Trognon 'stump,
butt end' was a "nom ridicule" and would be "un handicap," Trog-
nons throughout France rose up in their defense (Time, 11 Sep-
tember 1972, pp. 28-31). The need for variety impelled the Swedish
government to draw up a list of 900,000 unique new surnames by
having a computer randomly combine syllables (Landau 1967:16).
The problem that the computer was intended to solve is that there
are too few Scandinavian family names because many of them
were originally patronyms, limited in variety by the small number
of popular masculine given names on which they were based. Con-
sequently a few names like Olson, Hanson, Gustafson, or Swenson
occur with confusing frequency, even more so than comparable
names like Thompson, Johnson, and Jones in British and American
society. The convenience of the state is served by distinctness of its
Government efforts to control names have been widespread (Ren-
nick 1970; Wolf-Rottkay 1971:238), although not always successful.
They are, moreover, not limited to personal names; indeed in the
United States, where there is relatively little governmental control
of a person's name, official agencies are more likely to be concerned
with place names. Long-established names of topographical fea-
tures can be changed by government decree, as that of Cape
Canaveral was to Cape Kennedy, although a successful change re-
quires popular sympathy in support of it. When the federal Bureau
of Land Management decided to change the name of Whorehouse
Meadows, near Ontario, Oregon, originally named in honor of four
working girls who entertained the sheepherders (according to a
UPI dispatch in the Birmingham, Alabama, Post Herald, 4 Decem-
ber 1972, p. 3/2, a reference kindly supplied by James B. Mc-
Millan), local pride was mobilized and the Oregon state board for
geographical names petitioned for a return to the traditional desig-
nation. Official sanction or disapproval can create a legal class of
names or nonnames, but it does not necessarily affect popular nam-
ing practices, and it is with the latter rather than with official lists
that this study is concerned.
In logic there are special uses of the term name that have little
DEFINING THE TERM Name
or nothing to do with its use in onomastics. It may be used, for
example, in the sense 'that which names,' thus implying the ex-
istence of a named object (Quine 1960:180). In this use, any puta-
tive name for which there is no corresponding object cannot in fact
be a name. In a different use, name or proper name may be alter-
native expressions for a singular term, by which use proper names
would encompass not only Sir Walter Scott and Chomolungma,
but also the author of Waverley and the highest mountain in the
world (Church 1956:3-9).
Bertrand Russell made several important and influential state-
ments about proper names (1918; 1948:72-84), which have in com-
mon that they characterize a special technical use of the term
rather than describe what might be thought of as proper names in
ordinary language. For Russell, proper names in a "logically per-
fect language" are words for particulars-for qualities and com-
plexes of "compresent" qualities-the closest thing in ordinary
language being words like this or that. Words like Socrates are not
proper names at all, but "are really abbreviations for descriptions";
and therefore "proper names in the ordinary sense, if this is right,
are misleading, and embody a false metaphysic," being the "ghosts
of substances." Russell was careful to specify that his concern was
a technical use of the term name and not the ordinary one; yet his
statements about what ordinary names "really" are, if his technical
use of the term "is right," certainly suggest the transcendence of an
arbitrarily stipulated definition and an encroachment upon linguis-
tic description. Russell seems to be exhibiting the Humpty-Dumpty
syndrome, by which a definition that is first prescribed for a stip-
"ulated use is illegitimately extended to domains other than that for
which it was defined and ends by masquerading as a description
or a critique of actual use:
"There's glory for you!"
"I don't know what you mean by 'glory,' Alice said.
Humpty Dumpty smiled contemptuously. "Of course you
don't-till I tell you. I meant 'there's a nice knock-down
argument for you!'"
"But 'glory' doesn't mean 'a nice knock-down argument,'"
"When I use a word," Humpty Dumpty said, in rather a
scornful tone, "it means just what I choose it to mean-
neither more or less."
ON DEFINING THE PROPER NAME
"The question is," said Alice "whether you can make words
mean so many different things."
"The question is," said Humpty Dumpty, "which is to be
master-that's all." (Carroll 1871:268-69)
Humpty Dumpty, the arch-egghead, was playing a game accord-
ing to his own idiosyncratic rules. Alice, in her commonsensical
way, assumed that everybody has to use the public rules. Opposite
Russell's Humpty Dumpty, Sir Alan Gardiner has filled the role of
Alice in his analysis of the logician's views, which he called "as-
suredly the most fantastic theory of proper names that has ever
come to birth" (1954:57). Sir Alan was, to be sure, aware that Rus-
sell was talking logic and not onomastics, but inasmuch as Russell
himself blurred the distinction, the Alicean response is not wholly
inappropriate. Logically proper names, as Russell has characterized
them, simply do not exist in ordinary language. In terms of Straw-
son's (1959:9) distinction between descriptive metaphysics, which
"is content to describe the actual structure of our thought about the
world," and revisionary metaphysics, which "is concerned to pro-
duce a better structure," it is clear that Russell is a revisionist.
Linguistics, of which onomastics is a subdiscipline, has much to
learn from other studies like philosophy that deal to some extent
with the same subject matter; but because all linguistics is primarily
a descriptive study and only secondarily a revisionary one and be-
cause it approaches the subject with its own assumptions and aims,
it maintains its autonomy from other language-oriented disciplines
(Wells 1954). With the special uses of the term proper name in
logic and other branches of philosophy, this study will not be
Stipulated definitions are not, however, restricted to logicians.
Br0ndal's study (1948:92) of the parts of speech, which aims at a
set of definitions universally applicable, proposes an a priori con-
cept of proper noun that excludes from the class all morphemically
complex items like Eiffel Tower, New-port, Eng-land, Cam-bridge,
and even Ital-y and Turk-ey. It likewise excludes words that are
homonymous with common appellatives, such as Grace, Smith, or
presumably the Wash. From the standpoint of any ordinary concept
of name, these exclusions are arbitrary and unwarranted. For
Br0ndal (1948:145), an example of the proper name par excellence
is French ce: "Le frangais possede aussi un mot qui designe ex-
DEFINING THE TERM Name
clusivement un objet determine (R), concept qui definit le nom
propre. Ce nom propre universal est ce, ancien pronom d'monstra-
tif (RD) qui a perdu toute trace d'indetermination (D)." The
notion that French ce is a pure or universal example of a proper
name has some affinity with Russell's view that the closest thing
normal language has to logically proper names is a word like this
or that. The view is, however, no more congenial to descriptive
study when advanced by a Danish linguist than when advanced by
an English philosopher.
In general the concern of this study will be linguistic forms that
can be called proper names in ordinary language, rather than in any
of the technical or restricted uses of the kinds mentioned above.
Moreover, its concern will be the ordinary use of ordinary language
rather than out-of-the-way or idiosyncratic uses (Ryle 1953:108-
12). The emphasis on "ordinary language" should not, however, be
taken to imply either that this study is based on the Ordinary
Language school of philosophy or that it will investigate all the
ways the term name is ordinarily used. The first would be too re-
strictive since Ordinary Language philosophy is still a kind of
philosophy, not a kind of grammar; the second would be too loose
because name is freely used in a great many divergent ways in
everyday language. The center of study will be those things that
most grammarians and most students of onomastics have agreed to
call names. The purpose of the study is to survey some ways of
characterizing that class of things and of defining the term name
for onomastic purposes. In doing so, a major concern will be the
distinction between universals of naming and the facts of a par-
ticular language. Language-specific facts about names are interest-
ing and important for the study of the individual languages, but
linguistic universals of naming are what must be the ultimate con-
cern of any study of onomastics. The focus of this study is on proper
names as they are used in English, but a pervasive question is
whether the English treatment of names is idiosyncratic or em-
bodies a general fact about naming practice.
It is commonly accepted that there are universals of naming-
specifically, that all languages provide for a class of items that can
be called proper names, that in all cultures there are events that
are identified as "namings" (Pulgram 1954:3; Hockett 1958:311),
and further that names are a kind of noun in any language with a
class of words that can be called nouns (presumably any language
ON DEFINING THE PROPER NAME
whatever). Those views have not gone unchallenged. Thus Hamp
(1956:347) suggests that "it is probably only our typical West-
European thought patterns that lead us to confuse, or single out,
noun as the partner for name; if we were Semites or Algonquians,
we might well seek to distinguish names from predications or
verbs." Hacking (1968) has argued that there can be and very
probably are natural languages without particulars-among which
we must presumably include proper names-citing studies by
American anthropological linguists of Kwakiutl and Nootka, which
are supposed to have strikingly un-European structures.
Whether names are to be considered a kind of noun, and there-
fore to be contrasted with other kinds of nouns, may depend on
whether we have in mind their syntax, that is, their external gram-
matical relations with other items in some larger construction, or
their morphology, that is, the internal grammar of their own struc-
ture. That names, grammatically considered, function like nouns
within the sentence and are thus syntactically nouns is very likely
to be a universal truth. That names are made up according to the
same patterns that nouns are is, on the contrary, a very uncertain
universal. Indeed, in chapter 3 it will be seen that the second gen-
eralization does not hold even for English, so there is no need to
look to exotica to prove it wrong. Names are not always morpholog-
ically nouns (although it would also be possible to argue that nouns
are not always morphologically nouns either). However, that fact
still leaves open the validity of the weaker assumption that "not
only name-bearing, but also namebuilding is a universal human
practice, with the same elementary rules everywhere, just as human
language is basically the same physical and nervous performance
of human speech production and comprehension, regardless of the
multitude and variety of languages" (Pulgram 1954:20).
Whether a natural human language without terms for particulars
in its surface structure-that is, a natural language with the form
of the predicate calculus-exists or might be expected to exist is less
certain than Hacking would have it. Examples from American
Indian languages like Kwakiutl and Nootka have long, but uncon-
vincingly, been offered as proof that "Standard Average European"
(as Whorf called it) and Amerindian tongues present radically
different views of the world and everything in it. The problem is
that to compare the Weltanschauung of the two languages, one
must be paraphrased into the other, and paraphrases are tricky
DEFINING THE TERM Name
things. Quite ordinary expressions in quite ordinary languages, in-
cluding English itself, can be paraphrased in a way that would
make them seem very un-European, and exotic tongues like
Kwakiutl can be translated in a way that makes them the very
Doppelginger of any Indo-European language. Paraphrases lack
conviction. What is needed is a native command of both languages
coupled with the linguistic sophistication to analyze, contrast, and
interpret the two systems. So far such a happy conjunction of com-
petence and ability has not enfleshed itself in many persons. And
until it does, some caution is in order in accepting an analysis of
exotic languages that shows them to depart from what, on other
grounds, we might reasonably assume to be universal features of
Even though naming is in some sense "instinctual," there is no
reason to expect we can ever identify the "instinct" that underlies it.
Efforts to do so, like that of Feldman (1959), who believes names
to be an expression of antagonism toward the named object and
the use of names to be an aggressive act, are examples of psycho-
analysis gone to seed. It is, however, not too much to hope that
some universally valid criteria for defining names can be found. If
such criteria suggest an explanation for the universality, so much
the better. But having no explanation is better than having a
The fact that names are universal in human language (assuming
it to be a fact) need not imply any phylogenetic priority for them.
The view popular among some writers (Jespersen 1922:438; Chris-
tophersen 1939:64; Pulgram 1954:6) that proper names were the
oldest kind of noun or even the original kind of word is probably
due to an uncritical acceptance of a romantic view of the savage as
a simple, uncomplex soul, given to concrete thinking rather than
abstraction and preoccupied with the immediate present and sense
perceptions. The myth of the simple savage has also given rise to
fables about languages that consist of no more than two hundred
words, which are on a par with the just-so fable of some dawn-man
wandering through the primeval forest bestowing proper names on
the animals, trees, and rocks, and only gradually generalizing those
names to stand for any similar object, thereby discovering the con-
cept of species. It is, of course, beside the point that the Greek
term onoma kyrion, from which our proper name ultimately derives,
seems originally to have meant 'name, properly speaking, properly
ON DEFINING THE PROPER NAME
so called, in the proper sense of the word,' thus leaving common
names as second-class members of the part of speech. Despite its
etymology, etymology provides no insight into the "true" or "real"
meaning of words. What the Greeks thought about proper names
is an important part of the history of such words, but it tells noth-
ing about their origin, for Delphi was a long way from the dawn-
forest, and Chrysippus the Stoic, who is reputed to have originated
the grammatical term, was many a generation removed from the
The romantic view, which long antedates its twentieth-century
proponents cited above, was challenged before them by Max Miiller
(1891:512-19), who held the opposite view, namely that language
originated in general ideas expressed by general terms and only
later were some of them individualized and specialized into proper
names. If we are forced to choose sides in this dichotomy, it can
be said in support of Max Miiller that the way children learn
language seems to offer more support for his view than for the ro-
mantic one. Children are said to generalize through recognizing
likenesses more readily than to particularize through differentia-
tion. If ontogeny recapitulates linguistic phylogeny, we might
suppose with Max Miiller that languages began with something like
the roots he favored, expressive of general notions. However, he
himself seems to have suspected the dichotomy, for he cites Sir
William Hamilton's Lectures on Metaphysics to the effect that chil-
dren, and presumably also dawn-man, at first command neither
general terms nor individual ones, in our sense, but rather a single
undifferentiated sort of term, which seems vague and confused to
us, but out of which eventually develop the categories we expect.
There is much to recommend this view, in which neither the proper
nor the common name has any natural priority, historical or gram-
matical, over the other. Being correlative notions, it is not until
proper names and appellatives both exist that there can be any
contrast between them.
Universals of naming, whatever they may be, are necessarily re-
flected in the structure and use of particular languages, and so it is
with the latter that study must begin. A number of criteria have
been suggested for proper names in English, among which these
are the most prominent:
Proper names are capitalized.
DEFINING THE TERM Name
Proper names have no plural forms.
Proper names are used without articles.
Proper names do not accept restrictive
Proper names refer to single unique
Proper names do not impute any quali-
ties to the objects designated and are
Proper names have a distinctive form of
definition that includes a citation of
In the following chapters, all of the foregoing criteria, and some
additional related ones, will be examined for their appropriateness
to the way proper name is ordinarily used, for their mutual com-
patibility, and for their universality. In the process it will become
clear that what they define is not a single, natural class of words,
but rather a number of different sorts of names only roughly coter-
minous with one another. To distinguish these items from one an-
other, the following terminological distinctions will be made: on
the orthographic level, there are ORTHOGRAPHIC NAMES or CAPITAL-
IZED WORDS versus UNCAPITALIZED WORDS; on the morphosyntactic
level, there are PROPER NOUNS versus COMMON NOUNS; on the refer-
ential level, there are SINGULAR TERMS versus GENERAL TERMS; on
the semantic level, there are PROPER NAMES or simply NAMES versus
COMMON NAMES or APPELLATIVES. To talk about a "name" in the
abstract is to run the risk of hypostatizing a linguistic fiction. What
names are depends on whether we are concerned with semantics,
reference, grammar, phonology, or orthography. It depends also on
whether we are concerned with the universals of naming or with
the facts of a particular language.
Orthographical and Phonological Names
ON WHAT is the most accessible level of language,
the expression plane, it is conceivable that names
might be characteristically marked in any of the substances by
which language is expressed, most particularly in writing or speech.
And, indeed, capitalization in English writing is often cited as a
mark of the proper name. In the standard orthography there is a
class of written words that regularly begin with a capital letter,
regardless of the position they happen to occupy within the sen-
tence. This class of orthographic names (as they may be called)
correlates roughly with the class of proper names defined in other
ways, but the correlation is not exact.
Automobile names like Ford or Volkswagen are capitalized in all
uses. When they are names of the companies, they are grammati-
cally proper nouns because they are regularly used without articles,
but when they refer to the company's product-an automobile-
they are grammatically common since they then require an article.
Compare "Onassis sold the Ford and bought a Volkswagen" with
"Onassis sold Ford and bought Volkswagen." Product names like a
Ford, a Volkswagen, a Kodak, and a Frigidaire, are grammatically
common by every morphosyntactic criterion that has been proposed
to distinguish proper from common nouns, including the three
mentioned at the end of chapter 1. They are freely pluralized:
"Fords are selling well this year"; "He owns two Fords." They are
used with articles, and indeed require articles under exactly the
same conditions as any common noun: "He has a Ford/car, but the
Ford/car is second-hand," not *"He has Ford/car, but Ford/car is
second-hand." They accept restrictive modifiers without difficulty:
"The last Ford he bought was a lemon." The fact that the words in
question are trademarks is, as has already been observed, not
relevant to their linguistic status, although the fact that they are
ORTHOGRAPHICAL AND PHONOLOGICAL NAMES
grammatically common nouns may be a contributory factor to the
ease with which trademarks pass into generic use when other con-
ditions favor such a shift. The fact that a word is capitalized is no
guide to its status as a proper name on other levels.
Neither is the lack of capitalization a reliable clue to a word's
grammar. Thus, the word base in the expression first base, second
base, third base is always lowercase. Grammatically and semantic-
ally, however, it has two strikingly different uses. One use is for
any of the four stations at the corners of the infield, in which case
the term is grammatically common according to the criteria already
mentioned: "a base, bases, the base he is on." The other use is for
a particular one of the three stations other than home plate, in
which case the term is grammatically proper, rejecting the definite
article, as in "The runner was out at second (base)," for which
*"The runner was out at the second (base)" would be at best odd
and unidiomatic. The contrast between these two uses of base can
be seen in sentences like "First base is the first base a runner
touches," or "The first bases he stole were third base and second
base." Other items like first base and second base are top banana,
second tenor, south and dummy in bridge, and fullback as in "A
quarterback rarely plays fullback." Such terms are not often thought
of as proper nouns, chiefly perhaps because they are not capitalized,
but they satisfy the grammatical criteria for properness and there-
by demonstrate that some grammatically proper words are written
in lower case.
The examples so far cited show that items can be treated alike on
the orthographic level, either capitalized or lower-cased, although
they belong to different grammatical and semantic classes. The fol-
lowing examples will show that some items belonging to the same
grammatical or semantic classes are treated in different ways on
the orthographic level. The names of the planets, Mars, Venus,
Pluto, and so on, are always capitalized, whereas the name of the
third planet, earth, is usually spelled with an initial lower case,
although it would seem to be semantically as much a proper name
as the others. To be sure, the planet name Earth may be capitalized,
but it also may not, whereas Mars, Venus, and so forth would be
quite unusual with lowercase spellings. Similarly Caucasian, a noun
indicative of race, is regularly capitalized, whereas the noun white,
in the same racial sense, is regularly lowercase. Both words are
grammatically common nouns.
ON DEFINING THE PROPER NAME
Caucasian is typical of grammatically common nouns synchron-
ically derived from proper names, in this case from Caucasus or
Caucasia. Other examples are Parisian from Paris, Chinese from
China, Trotskyite from Trotsky, and Michigander from Michigan.
Such words are proper orthographically but not syntactically. For
some words, capitalization serves incidentally to distinguish hom-
onyms: Democrat versus democrat, Shaker versus shaker, and Odd
Fellow versus odd fellow. The orthographic contrast with homonyms
is, however, neither diachronic explanation nor synchronic justifi-
cation for their capitalization, which is to be found rather in their
connection with proper names like the Democratic Party and the
Independent Order of Odd Fellows. Finally, some words are spo-
radically capitalized honors causa: "a Committee meeting," "a
future Convention," "the present Administration," "Transforma-
tional-Generative grammar," and recently Black as a racial term.
There is a good deal of variation in the capitalization of some of
the words cited above, but that variation is itself evidence of the
poor correlation of orthography with grammatical or semantic facts.
The Chicago Manual of Style (1969:147-94) devotes nearly fifty
pages to rules for capitalization, chiefly in an effort to resolve the
problematical cases, and actual unedited usage shows a great deal
more variation than style guides suggest. Variation exists partly
because the rules for capitalization are sui generis, rather than
reflections of systematic distinctions made on other levels of the
Grammarians often assume a greater correlation between capital-
ization and grammatically proper nouns than in fact exists. Thus
Pulgram (1954:37), in one of the more extreme statements of
the kind, urged his readers to "notice the prevalent spelling prac-
tice in English today: names of apples with majuscules-they are
strictly proper names to the great majority of speakers; names of
automobiles with majuscules-they, also, are proper names, associ-
ated with the name of the producers or the factory, they are trade-
marks; but the names of breeds of dogs are mostly spelled with
minuscules-because they have, in general usage, long ago ceased
to be proper names and have become, through their importance
to humans, familiar general nouns." He goes on to conclude that "in
modem English the spelling actually is indicative of the usage of
a word," but his argument might more precisely be said to be that
the spelling determines whether a word is proper or common, for
ORTHOGRAPHICAL AND PHONOLOGICAL NAMES
only on that basis is it possible to make sense of the idea that
Delicious and Datsun are proper whereas dachshund is common.
Not many grammarians who have been influenced by American
structuralism, with its insistence that language is speech and that
writing is at best a derivative representation, would subscribe to
the naive assumption that what is capitalized is a proper name;
but the opposite assumption, that majuscules are used for words
that can be shown on syntactic or semantic grounds to be proper,
has been made more or less explicitly. Thus Long (1969:109) sug-
gests that "the honorific capital letters with which most proper
names are begun are a tribute the written language pays to indi-
viduality," and Paul Roberts in one of his high school textbooks
(1967:17) cautions the student, "Obviously we can't find out which
nouns are proper ones by just noting which nouns are capitalized.
.. We should still have to ask ourselves how [the writer] knows
which nouns to capitalize and which not." The assumption is that
we can define proper nouns some other way-by their behavior with
articles, for example-and that all nouns so defined will be capital-
ized. There may be pedagogical justification for such an assumption,
made also by Quirk et al. (1972:160, 164), but it is not wholly
accurate as a description of linguistic fact.
Gleason (1965:186) has made the valid and significant point that
because the use of capitals in spelling has no phonological reference
whatever, insofar as capitalization represents anything from another
part of the linguistic system, it is grammatical information about
the word class of proper nouns. But capitalization turns out to give
relatively little information about any other level of language, for
the orthography is an autonomous system with its own, often idio-
syncratic, rules and has only an approximate correlation with syn-
tax or anything else. Most but not all proper nouns are capitalized
in English, and a great many things that certainly are not proper
nouns are regularly capitalized. Present-day English has some
words like Chevrolet that are usually capitalized, some like Roman
that are often capitalized, some like devil that are occasionally
capitalized, and some like first base that are rarely capitalized.
Among the several reasons for a majuscule is that a word is thought
to be "proper" in some way other than orthographical, although
forms like first base and earth show that there are grammatically or
semantically proper words that are written in minuscules.
It is also worth noting that the Modem English use of capitals
ON DEFINING THE PROPER NAME
is limited to a single etat de langue. Proof of that obvious fact is
that classical writing used only majuscules (or rather, that the
minuscule-majuscule distinction did not appear until early medieval
times) and that the post-classical use of majuscules has varied
greatly. Thus Modern German begins all nouns with capitals where-
as Old English used capitals only to begin sentences. German spells
Mann with a capital, and Old English spelled Alfred all lowercase.
In neither system does the use of majuscules define a class one can
identify even approximately as proper names. Furthermore, other
periods, other languages, and other writing systems have their own
devices for identifying orthographic names. Thus, in the eighteenth
century, proper names were often italicized. In the Initial Teaching
Alphabet, the first letter is boldface. In the Shavian alphabet,
proper names are preceded by a raised dot, and in the alphabet of
the International Phonetic Association, by an asterisk. In hiero-
glyphic Egyptian, the cartouche set off royal names. In cuneiform,
divine names were accompanied by a special identifying symbol as
a determinative. And doubtless other devices have been used. Two
things, however, are clear: written representation of names is a
language-specific and highly variable matter; and the class of writ-
ten names in present-day English is not the same as the class of
grammatical or semantic names.
Indeed, the fact that English has any kind of written symboliza-
tion for "nameness" is remarkable, as Gleason has pointed out, be-
cause generally writing is more restricted in what it can signal than
is speech; and English has no phonologically defined class of names,
nor is it likely that many languages do. Still, it is conceivable that
a language might indicate names by their sound. One artificial
language, Loglan (developed by James Brown at the University of
Florida in the 1960s), has proper names that can be recognized
from their pronunciation, because they and they alone are poly-
syllabic words ending in a consonant. But in natural languages such
phenomena must surely be rare. It has been reported by Harms
(1968:13) that Finnish proper names are exempt from certain
phonological rules affecting all common nouns. In that case there
would seem to be a phonologically defined class of words that
coincides at least partly with the class of semantic proper names,
but it is doubtful that the correlation is exact.
It is, to be sure, well known that names often have a form at
variance with what the general rules of historical sound change
ORTHOGRAPHICAL AND PHONOLOGICAL NAMES
would lead us to expect from their etyma. Instances like York from
Eoforwic, Wooster from Worcester, and maudlin from Magdalene
are well known, but more parochial examples, like St. Louis' Forest
Park, whose first word is locally pronounced to rime with bars, are
not lacking. The spelling of proper names is just as great a problem;
it is relatively easy to produce outrageous examples-Cholmondeley
and Marjoribanks are familiar ones cited by Krapp (1925:169), and
Scots Auchinleck and Ruthven, pronounced in ways that might be
spelled Affleck and Rivven, are other favorites. In these respects,
however, names do not differ from equally well-known common
words like boatswain, forecastle, woman (from wifmann), lord
(from hlafweard), or even one, whose initial /w/ is unexpected.
The unpredictable phonological development of such words and
their often idiosyncratic spellings in present-day English are re-
lated phenomena, one being a consequence of the other. Such real
or apparent exceptions can be found among both proper and
common words, so Br0ndal (1948:92) was indulging in hyperbole
when he remarked that the pronunciation and spelling of personal
names is "sans droit ni loi." Utley (1963:163), on the other hand,
may have been too optimistic in his belief that "a whole new gram-
mar of sound-change can develop from the close study of names
and their startling variants," unless we are content with a historical
grammar in which each word has its own history. The resistance
of proper names to the historical changes that affect a language
may have produced the situation in Finnish wherein proper names
are exempt from synchronic phonological rules. Such a rough
characterization is, however, no definition. Moreover, a phono-
logically defined class of names, if it occurs in any language at all,
would certainly be an oddity of rare occurrence and language-
Phonologically signaled proper names are either nonoccurring
or extremely rare. Orthographically defined classes of words that
correlate roughly with names as they are defined on other linguistic
levels exist in some languages, including English. But English cap-
italized words are an autonomous class that have to be defined on
the level of expression in orthographic terms; they reflect distinc-
tions on other levels only fragmentarily and inconsistently.
RAMMARIANS have given a good deal of attention to
the class of proper nouns defined in morphosyn-
tactic terms. Syntactic criteria-that is, the item's relationships to
other items in the same construction-have been most popular;
but some attention has been given also to morphological criteria-
that is, the relationships between the parts that make up the item
itself when it is more than a single morpheme. The syntactic cri-
teria most often cited are that proper nouns (1) are not used in
the plural, (2) are not used with the range of articles available
to common nouns, and (3) do not accept restrictive relative clauses
as modifiers. The first criterion has been used by Jespersen (1924:
69), Bloomfield (1933:205), S0rensen (1958:156), Long (1961:39,
1969:109), and Strang (1968:114); the second by Jespersen (1909-
49:7.437, 1924:69), Poutsma (1914:2.564), Kruisinga (1932:3.344),
Bloomfield (1933:205), Christophersen (1939:79), McMillan (1949:
243), Gardiner (1954:21), Hockett (1958:311-13), S0rensen (1958:
156), Sledd (1959:244), Long (1961:39, 503, 1969:109), Chom-
sky (1965:100), Gleason (1965:134), Roberts (1967:17), Strang
(1968:114), Zandvoort (1969:121), and Quirk et al. (1972:128,
160); and the third by Jespersen (1909-49:7.472), S0rensen (1958:
156), Utley (1963:173), Smith (1964:38), Long (1969:109), and
Any of the putative criteria taken alone is, however, inadequate
for a satisfactory definition of proper names. Thus all of the follow-
ing are possible:
I know a girl.
I know a Raquel.
All the girls I know are buxom.
All the Raquels I know are buxom.
A word like Raquel can be used in the singular or the plural, with
a or the, and with or without restrictive modifiers. There are two
probable senses of the word in such uses: 'person bearing the
proper name in question' and 'person with some characteristics
reminiscent of a known person bearing the name in question.' The
second sense may seem less proper than the first; when, for a given
name such as quisling or martinet, it ceases to be merely a nonce
use and is conventionalized so that the name stands directly for
the characteristic, 'traitorous puppet of foreign invaders' or 'strict
disciplinarian,' rather than for the historical person, the name has
clearly been commonized. There is, however, no clear reason to
deny that the first sense is a proper use of the word. To anticipate
chapter 6, if Raquel in "I know a Raquer' can be explained as 'a
person with the name Raquel,' the name in "I know Raquel" can
likewise be explained as 'the person with the name Raquel.'
The existence of the sentences cited above with the noun Raquel
is an embarrassment for those who would maintain that proper
nouns are used only in the singular and without articles or re-
strictive modifiers. Two approaches have been taken to explain
away the offending sentences, both desperate measures. On the one
hand, it has been proposed, in what may be called the Doppel-
giinger Effect, that Raquel in the sentence "I know Raquel" and
Raquel in the expressions "a Raquel" or "all the Raquels I know"
are really different, albeit homonymous, words, one a proper noun
and the other a common noun (Bloomfield 1933:205; Strang 1968:
114; Chafe 1970:292; Hockett 1958:312 speaks of such names as
having been "deproper-ized"). If this explanation is accepted, it
follows that beside every proper noun in the language there is
such a homonymous common noun. The consequence is unaccept-
able because it treats as an isolated, unpredictable lexical peculiar-
ity what is really a general and completely predictable grammatical
fact. Given any proper noun in English-from Aaron to Zurich-
English speakers are free to talk about "an Aaron" and "the Zurichs
that never were." Whatever can be accounted for by a general rule
must be regarded as a grammatical fact, not a lexical accident
(Bloomfield 1933:274). It is consequently a grammatical fact that
all proper nouns can be used in the plural, with articles, and modi-
fied by restrictive clauses.
The other desperate solution, which might be called the Ostrich
Caper, makes use of the Saussurian dichotomy between langue and
ON DEFINING THE PROPER NAME
parole to maintain that the use of Raquel in expressions like "a
Raquel" and "all the Raquels I know" is an accident of the indi-
vidual speech act and not a fact of the general language system
(Gardiner 1954:12; Sampson 1970:131 is undecided whether the
use is competent or only part of performance). But the uses under
consideration are clearly a part of the English language because
they are used by all native speakers and are not the result of this
or that individual's caprice. To bury one's head in the langue will
not suffice to make the offensive expressions disappear.
Although the criteria most often put forth as defining the gram-
matical class of proper nouns are inadequate, it is nonetheless true
that there is such a class and that the commonly advanced criteria
are based on correct, if insufficiently detailed, observations. Clar-
ence Sloat (1969) has presented the most accurate characterization
of the English proper noun, concerning which he observes that
"the definite article will appear as zero before singular proper nouns,
except when it is heavily stressed or they are preceded by restric-
tive adjectives or are followed by restrictive relative clauses."
Sloat's observations need several qualifications, but they are basi-
Proper nouns may be indefinite, and when they are, they take
the full range of limiting modifiers available to common unit nouns.
(Whether proper nouns are ever mass nouns is a question that will
be taken up later in this chapter; it is clear that most are unit
nouns, and the question does not affect the argument.) Thus we
have sentences like the following:
I know a Raquel.
I know some Raquels.
I don't know any Raquels.
I know every Raquel.
When a proper noun is definite, the article the, or one of the
other definite determiners like this or that, is overtly present if the
noun is plural:
I know the Raquels.
I know those Raquels.
These Raquels are different from the others.
The article is also present if the definite proper noun is restricted:
I know the Raquel who has brown eyes.
I know the Raquel with brown eyes.
I know the brown-eyed Raquel.
Further, the article is present if it is the information focus (Halli-
day 1967-68:3.203) for the sentence and thus carries the sentence
I know the Raquel.
If, however, none of the four conditions obtain, so that the proper
noun is singular, unrestricted, and definite without the definiteness
being regarded as new information, the article cannot be present:
I know Raquel.
NOT *I know the Raquel.
Apart from this one relatively minor peculiarity, proper nouns
are syntactically like other unit nouns:
A man is here. A George is here.
Some men are here. Some Georges are here.
The man is here. 0 George is here.
The man you know is here. The George you know is here.
The men are here. The Georges are here.
This one irregularity in the definiteness paradigm is, however, what
defines the grammatical class of proper nouns. The three criteria
noted earlier are thus correct, in that, with the added qualification
concerning new information, their intersection is the irregularity
that sets proper nouns apart from common ones.
There are a number of exceptions or apparent exceptions to the
foregoing generalization. A large group of words like Volkswagen,
Spaniard, and Delicious (apple) are regularly capitalized and
therefore sometimes called proper, but they require the in circum-
stances that preclude the definite article for proper nouns: "The
Spaniard in the Volkswagen wanted the Delicious rather than the
ON DEFINING THE PROPER NAME
Winesap." These words offer no real problem. They are common
nouns that, for a variety of reasons having nothing to do with
grammar, are capitalized. They were discussed in chapter 2.
Words that are indeed proper names by the semantic criterion
to be considered in chapter 6 include some that, like the preceding
group, are syntactically common nouns and others whose syntax
is that of proper nouns. These words fall into fairly clear subclasses
that can be illustrated by the following names: Vesuvius, Mount
Olivet, Stone Mountain, the Mount of Olives, the Rocky Mountains,
and the Matterhorn.
Words like Vesuvius are straightforwardly names that are also
proper nouns: Bermuda, Kalamazoo, Broadway, Carl, and Thomp-
son are other examples. Words like heaven, hell, purgatory, limbo,
and paradise, which are sometimes capitalized and sometimes not,
are proper nouns in their syntax, but can be defined in a way other
than the definition schema for proper names to be developed in
chapter 6. Jespersen (1909-49:7.577) calls them quasi-proper
names. There is also a group of nouns that are either proper or
common, for example parliament, which is proper in "Parliament is
meeting" but common in "The parliament is meeting." It is not
clear that there is any semantic contrast between the two gram-
matical uses. Similar items are congress, earth, scripture, god,
teacher, cook, sister, father, and other words of occupation and
family relationship. Contrasting grammatical use of some of these
words implies semantic differences, at least in the form of pre-
suppositions. Thus "God will help" implies the speaker is a mono-
theist, whereas "The God will help" implies he is a polytheist.
"Teacher will help" and "Sister will help" imply that the person
referred to has the designated relationship to either the speaker
or the one spoken to. The determiner-less noun may be understood
as elliptical for either "my teacher" or "your teacher," but not
usually for "his teacher." If A says to B about C, "Brother is here,"
an eavesdropper can confidently assume that C is brother to either
(or both) A or B. If he is not, the appropriate form is "The brother
is here," which has its own implication, namely that C is a brother
to someone other than A or B.
Some words that are quite clearly semantic appellatives (as they
will be defined in chapter 6) are grammatically proper nouns, for
example zero as a point on the temperature scale (unlike the freez-
ing point, which is semantically similar, although a common noun)
and sea level as a point on the altitude scale (unlike the fall line).
First base and other items of the kind have been mentioned in
chapter 2. Another similar set (pointed out to me by James B. Mc-
Millan) includes librarian, as in "John is (the) librarian at Chatta-
hoochee," and so also headmaster, dean, chairman, sheriff, janitor,
purser, cashier, manager, conductor, pilot, and a good many others,
but not stenographer, student, sophomore, plumber, dentist, cabbie,
stewardess, and so on. Those appellatives that can be used as
grammatically proper nouns denote positions often involving some
authority over others and, in grammatically proper use, denote
one-of-a-kind positions within their hierarchies; such appellatives
also have more frequent use as titles and are more often capitalized
than the appellatives that are strictly common nouns in their gram-
mar. To say "John is librarian" implies that he is either the only
or the head librarian in a position of hierarchical responsibility. To
say *"John is dentist" would be odd because dentists do not have
a hierarchically defined position.
Not every use of a noun without the is proper use, however. To
be proper, a noun must be regularly used without the under the
appropriate circumstances. Thus, the-less uses of school, college,
prison, court, market, church, home, bed, breakfast, lunch, supper,
automobile, plane, ship, train, and so forth are not enough to make
of those words proper nouns. When they are the objects of prepo-
sitions they may occur without the, as in "to school," "in bed," "at
breakfast," "by automobile," and also in other uses in which they
denote a kind of activity: "College is part of an education." In some
such uses, the words are mass nouns, as can be seen from the
mass-noun modifiers that occur with them: "enough school, too
much school." In definite use as unit nouns, they require the: "He's
going to the school."
Proper names of the type Mount Olivet or Stone Mountain con-
sist of a general term and a specific, the order depending on what
the general term is. Most general terms come second: Yale Univer-
sity, Sherwood Forest, London Bridge, Biscayne Bay, Pennsylvania
Avenue, Staten Island, Carson City, Niagara Falls, Carnegie Hall,
Lincoln Center, and Penn Station. The order general-specific occurs
with fewer general terms: Cape Hatteras, Fort Sumter, and Lake
Leman (compare the less common order for lake names in Creve
Coeur Lake). Such composite names are similar to combinations
like chanteuse Hildegard, pianist Liberace, comedian Jack Benny,
ON DEFINING THE PROPER NAME
legislator Bella Abzug; and these in turn blend into combinations
of titles and names: Senator Smith, Mayor Daley, Doctor Rubin,
and Professor Hill. Apart from the obvious semantic contrast of
place versus personal names, the chief difference among these com-
binations is the extent to which the general term has been institu-
tionalized as an accompaniment to the name.
In addition to the foregoing combinations of a general and a
specific term by juxtaposition, many names are of the type the
Mount of Olives, a noun phrase with the general term as head and
the specific as the object of a modifying prepositional phrase: the
University of Wisconsin, the Forest of Arden, the Bridge of Sighs,
the Bay of Biscay, the Avenue of the Americas, the Isle of Man,
the City of Chicago, the Falls of the Rhine, the Cape of Good Hope,
the Castle of Otranto, the Lake of Geneva, and the Sea of Japan.
Despite the semantic parallelism between names like the Mount of
Olives and Mount Olivet, the former is grammatically common and
the latter grammatically proper, since the first but not the second
requires the under those circumstances in which it is required for
Expressions of the type the Rocky Mountains are also examples
of proper names that are grammatically common nouns. They too
are very frequent: the Great Plains, the Ozark Plateau, the Grand
Canyon, the Bering Strait, the Persian Gulf, the Brooklyn Bridge,
the Canal Zone, the Ivory Coast, the Orange Free State, the Soviet
Union, the Black Forest, the China Sea, the Okefenokee Swamp,
the Gobi Desert, the Pacific Ocean, and the Suez Canal. The normal
pattern is illustrated by the Mississippi River, in which the specific,
Mississippi, serves as an adjunct within a noun phrase, the river,
with the general term as head. More rarely the specific follows the
general term, as in the River Thames. Among these constructions
also, there is a lack of parallelism on the semantic and grammatical
levels. The Great Salt Lake and Lake Michigan are names that are
respectively common and proper nouns, whereas the twelfth chap-
ter and chapter twelve are appellatives with the same grammatical
difference between them.
Chapter twelve is an appellative because, as will be pointed out
in chapter 6, its application is predictable in ways that the applica-
tion of a name is not. A chapter that comes between chapters
eleven and thirteen will necessarily be called chapter twelve by a
general rule of numbering; there is no general rule to predict that
the lake situated between Lakes Huron and Superior will be called
Lake Michigan. Chapter twelve, however, is used without the under
the same circumstances as Lake Michigan, and therefore both are
grammatically proper nouns. The grammar of numbered items (as
James B. McMillan has pointed out to me) is complex. An optional
transformation might be postulated to relate surface common noun
phrases like the twelfth chapter to surface proper noun phrases like
chapter twelve, thus obliterating any distinction between them on
a deeper level of analysis. But the third anniversary does not have
a corresponding anniversary three in ordinary use, nor does the
eightieth highway relate simply to highway 80. Moreover, with some
nouns the ordinal form is syntactically proper: Eighth Avenue. Ex-
pressions such as aisle two, precinct seven, (the) second grade, the
third floor, uranium 235, P.S. 95, size twelve, vitamin A, section L,
Mathematics 201, and grade B or B grade (movies) present syn-
tactic and semantic problems that await exploration but are beyond
the scope of this study. Here it is enough to note that the members
of pairs like chapter twelve and the twelfth chapter or like Lake
Michigan and the Great Salt Lake differ in their surface grammar.
Another varied class of names are those like the Matterhorn,
consisting of only a specific preceded by the definite article: the
Ukraine, the Transvaal, the Maghreb, the Negev, the Crimea, the
Saar, the Tirol, the Congo, the Sudan, the Peloponnesus, the Helles-
pont, the Dardanelles, the Skaw, the Everglades, the Palisades, the
Lido, the Alhambra, the Kaaba, the Kremlin, the Pentagon, the
Alamo, the Escorial, the Vatican, the Hague, the Iliad, the Torah,
the Koran, the Dhammapada, the Cid. Despite the clear status of
such terms as semantic names, the presence of the marks them as
grammatically common. Whether a name requires the or does not
is largely unpredictable. James Fenimore Cooper was apparently
unable to decide whether the name of the title character of The
Deerslayer was a characterizing epithet and thus a common noun
that requires the or a genuine proper noun that rejects the. The last
two references to the novel's hero are these: "Fifteen years had
passed away ere it was in the power of the Deerslayer to revisit the
Glimmerglass" and "The heart of Deerslayer beat quick as he found
a ribbon of Judith's fluttering from a log." Although Deerslayer is as
much a name in one use as another, its grammar fluctuates between
that of common and proper nouns. Alternate names for the same
referent and names for similar referents also show that it is highly
ON DEFINING THE PROPER NAME
variable whether such names are proper or common nouns. Thus
the Caucasus is matched by Caucasia, the Argentine by Argentina,
the Netherlands by Holland, the Matterhorn by Vesuvius, the Bow-
ery by Harlem, and the Bastille by Alcatraz.
Some names of this sort have the idiomatically associated
with them (the Hague) and have had it since their importation
into English. Some have the as a result of ellipsis: the Bodleian
(Library) and the Metropolitan (Museum or Opera). Perhaps
ellipsis is also a reasonable explanation for the use of the with ship
names like the (ship) Mayflower, the (submarine) Nautilus, and
the (carrier) Forrestal. Others can hardly be explained as any kind
of ellipsis but seem to result from a specialization of general terms:
the Piedmont, the Narrows, the Downs, the Weald, the Wolds, the
Strand. Such names are similar to the use of any general term with
definite reference: the Channel (with reference to the English Chan-
nel regardless of what other channels may happen to be in the
proximity of the speaker), the City (central London or New York),
the Coast (California), the Rock (Gibraltar, Alcatraz, or even the
Prudential Insurance Company), the Lord, the Savior, the Father,
the devil, the sun, the moon, and the Government. It is difficult to
distinguish the foregoing from any common noun used with the
when the context makes the reference clear: "The president has
nominated a replacement for the chief justice of the court"; "The
president has appointed a replacement for the dean of the college."
If the context is adequate to identify the referent, we talk about the
kitchen, the pool, the river, the capitol, the constitution, the lab, the
chairman, the end zone, and so on, like the Harvard secretary who
is supposed to have informed a caller that the President was in
Washington visiting Mr. Taft. Such expressions are common nouns
in grammar and appellatives in meaning. The similarity between
definite common nouns like the queen and names like Elizabeth
has been noted (Christophersen 1939:70) but should not be exag-
gerated. The similarity lies in their mutual definiteness and thus
equivalence of reference. It is a confusion of categories that makes
for grammatical nonsense to say that the queen, with reference to
Elizabeth II, is a name or a proper noun.
It is sometimes suggested that many of the above terms and
others like the universe, the creation, and the zodiac constitute a
special group of uniquees" appellative terms denoting a class of
only one member (Christophersen 1939:24-25). Jespersen (1909-
49:7.482) has cogently criticized this view on the ground that in
language there are no such classes. The fact that some words occur
chiefly in the singular with the is an incidental consequence of the
way the world is, not a significant generalization about language.
Equator is used with the definite article the more than with the
indefinite article an or with other determiners like my because
there is, for each rotating body, just one equator; and by force of
circumstance the earth is the rotating body whose equator is most
often a topic of discourse. The fact that the equator is more com-
mon than an equator or my equator tells more about the world than
about the linguistic system. It is therefore misleading to set up a
class of "unique" nouns that generally occur in the singular with
It has also been proposed that the vocable the used with proper
nouns like the Hague or the Mississippi is not the definite article,
but something else: a part of the noun itself (Christophersen 1939:
61; Chomsky 1965:100; Roberts 1967:17; Strang 1968:114) or a
meaningless noise (S0rensen 1958:156). Such interpretations are
based on the assumption that names are not accompanied by deter-
miners freely chosen from the definiteness system. It has already
been argued that the assumption is mistaken. Proper names are
generally used in the definite form, just as they are generally used
in the singular and without modification, but they need not be so
used. To say that the in the Bronx is something other than the defi-
nite article will entail that the category of definiteness, of which
the Bronx is an exponent, is not explicitly realized in the name. This
interpretation will make the Bronx and Brooklyn alike, in that
neither uses the definite article to signal its definiteness, but the
parallelism is achieved at the cost of a reduction ad absurdum: al-
though the Bronx is definite, its the cannot be the definite article
but must instead be an empty morph, definiteness having zero
realization for all proper names. This triumph of rule over reality
is similar to one structural explanation for irregular forms like man/
men, which held that the vowel difference is a nondistinctive con-
sequence of a zero suffix on the second word. Instead of such occult
explanations, we can say (1) for most nouns that are singular un-
restricted names, the category of definiteness is unmarked; (2) for
some, however, it is marked in the usual way, by the; (3) nouns
with marked definiteness are "common," whereas those with un-
marked definiteness are "proper."
ON DEFINING THE PROPER NAME
The items so far discussed are only apparent exceptions to the
definition of proper noun given earlier in this chapter. That they
should ever have been thought exceptions is due to a confusion of
categories from different linguistic levels, chiefly of proper nouns
on the grammatical level with capitalized words on the orthographic
level and names on the semantic level. Quite simply, Vesuvius,
Parliament, noon, zero, Mount Olivet, and Stone Mountain are all
proper nouns because they meet the definition of that grammatical
class, whereas the Volkswagen, the parliament, the afternoon, the
freezing point, the Mount of Olives, the Rocky Mountains, the Mat-
terhorn, and the sun are all common nouns because they do not
meet that definition. Perhaps the earliest grammatical treatment of
proper nouns to recognize that cooccurrence with the definite ar-
ticle is the defining characteristic of the syntactic class and to accept
the consequences of that recognition was James B. McMillan's
study of place names (1949). If other grammarians had followed the
consistency of his grammatical description, they could have avoided
the dilemma that results from a monostratal view of language that
assumes names and proper nouns must be isomorphic. There is not
a perfect isomorphism between semantics and grammar. Mount
Olivet and the Mount of Olives are both semantically names, just
as zero and the freezing point are both semantically appellatives;
but the first member of each pair is grammatically a proper noun
and the second grammatically a common noun. Interlevel dis-
crepancies of this kind are rife in language, but cannot be well
understood without a view of linguistic structure that includes
There are, however, some genuine exceptions to the rule describ-
ing the use of the with proper nouns. First there is the use of the
definite article with a Scottish family name to denote the head of
the clan: the MacNab, the MacGregor. This semititle use is clearly
marginal, for not only is it limited to the names of one ethnic group
(the Humplemeyer and the Jones are nonoccurring), but most
English speakers do not in fact use it at all. The historical status
of the form is also marginal. The modern use seems to have been
an innovation of Sir Walter Scott (Dallas 1913) popularized by
his readers along with a number of words like glamor, which the
OED says was extended to the literary language by his novels.
There is another use of the before personal names that is also
marginal, although not so limited geographically or chronologically.
Christophersen (1939:114-15) cites Shakespeare, "My ancestor did
from the streets of Rome the Tarquin drive"; Carlyle, "Stout
Choiseul would discern in the Dubarry nothing but a wonderfully
dizened Scarlet-woman"; the Manchester Guardian, "The real farm-
er's boy of these days is a smart lad with a motor-bicycle and
views about the Dietrich and the Garbo"; and Dorothy Sayers,
"There's the Meteyard. I must tell her what Armstrong said." The
use, which seems faintly foreign and is reminiscent of French la as
in La Farge or German die in reference to female artists and enter-
tainers in the popular press, suggests preeminence of the person so
distinguished, either in earnest or with a faintly mocking tone. This
use is uncommon except perhaps in a somewhat old-fashioned ex-
pression like the Magdalene, which is analogous to the Virgin.
The two foregoing exceptional uses of the with proper nouns
that normally reject the article are real exceptions to the general
rule defining the grammatical class, but they are so limited in their
scope that the generalization is not seriously affected by them.
Another exception of a more serious nature pertains to the use of
restrictive modifiers with the-less proper nouns. The exact nature
of this exception depends on the kind of proper noun and on the
kind of restrictive modifier it has. The pattern that we would expect
is illustrated by the following paradigm:
George, who's governor the George who's governor
George, from Alabama the George from Alabama
conservative George the conservative George
If a proper noun is modified by a nonrestrictive element, as in the
first column, it is used without the; but if it is modified by a restric-
tive clause, phrase, or word, as in the second column, the is re-
quired. When the proper noun is a personal name and when the
modifier is a clause or phrase, the pattern is followed fairly well.
When, however, the modifier is an attributive adjective, two kinds
of exception are common. First, a restrictive adjective may be used
without the: "Are you voting for conservative George or liberal
George?" Since the function of the adjectives in this sentence is to
distinguish Georges, the adjectives are clearly restrictive. Second,
a nonrestrictive adjective may be used with the: "the irrepressible
Hubert." Because such a phrase can be freely used in a context
ON DEFINING THE PROPER NAME
involving only one Hubert, the adjective is clearly nonrestrictive. It
would seem, then, that the generalization describing the use of the
with proper nouns fails when the noun is modified by an attributive
There is, however, an important characteristic of restrictive ad-
jectives modifying personal names that allows the generalization to
be maintained in a slightly modified form. Consider the following
sentences, in which clause accent is marked, thus indicating what
is to be taken as "new information" in the sentence:
The young Yeats was talented, but the old Yeats was a
Y6ung Yeats was talented, but 61d Yeats was a genius.
Because clause accent is information-bearing, it is nonpredictable
(Bolinger 1972); consequently other positions for the accent are
possible in both sentences. Indeed, the two sentences can have ex-
actly the same accent pattern-but not with the same effect. The
accent positions indicated above are the normal, or unmarked, po-
sitions for the accent. Deviations from these positions imply special
contrast, emphasis, or the like. Thus, when the information focus is
on a part of the sentence other than the noun phrase, as in the first
example, the restrictive meaning of the modifier is signaled by the
presence of the article. When the information focus is on the re-
strictive adjective, however, the article can be omitted. Thus, to
the definition of proper nouns must be added the provision that the
definite article is also omitted optionally when the information focus
falls on a restrictive attributive adjective.
Nonrestrictive modification with the like "the irrepressible Hu-
bert" is a different problem. Such constructions are extremely
the peripatetic Kissinger
the admirable Crichton
the Venerable Bede
the Great Gatsby
the Reverend Arthur Dimmesdale
the Honorable Sam Massell
With these constructions it is instructive to compare postnominal
modifiers of the type "Alfred the Great," "Ethelred the Unready,"
"Ivan the Terrible," and noun appositives like "Priestley the chem-
ist," "the chemist Priestley"; "Mary the Virgin," "the Virgin Mary."
Common noun phrases that include the definite article can occur
either before or after a proper noun to which they are appositive.
Postnominal adjectives preceded by the, such as "Alfred the Great"
or "Hubert the irrepressible," are best accounted for as headless
noun phrases: "Alfred the Great (one)," "Hubert the irrepressible
(one)," thus making them parallel to noun appositives and suggest-
ing the proportion:
Priestley the chemist: the chemist Priestley::
Hubert the irrepressible: the irrepressible Hubert.
Given the foregoing proportion, we can say that just as "Hubert,
who is irrepressible" can be paraphrased as "irrepressible Hubert,"
so "Hubert, who is the irrepressible (one)" can be paraphrased as
either "Hubert, the irrepressible" or "the irrepressible Hubert."
Under this interpretation, constructions like "the irrepressible Hu-
bert," "the peripatetic Kissinger," and "the Venerable Bede" do not
violate the generalization that nouns modified nonrestrictively omit
the, because the article accompanies the adjective rather than the
noun, giving the constituent structure ((the irrepressible) Hubert)
as opposed to the constituent structure of noun phrases with re-
strictive adjectives: (the (conservative) George).
Thus far only proper nouns that are personal names have been
considered. Other kinds of proper nouns, such as place names, offer
somewhat different complications. Like personal names, place
names modified by a restrictive clause require the: "the America
that the colonists knew." Unlike personal names, however, place
names have the optionally when they are modified by restrictive
phrases: "(the) America of the colonial period"; and when the re-
strictive modifier is an adjective, the is usually impossible or un-
likely: "ancient Greece," "eastern Europe," "outer Mongolia," "cisal-
pine Gaul," "residential Chicago," "theatrical New York," "coastal
America," "contemporary London." Some adjectives can be used
either with or without the, although with some probability of a
contrast of meaning: Thus, "the old Atlanta" and "the new Israel"
may suggest a chronological dimension, 'the Atlanta of yesteryear'
ON DEFINING THE PROPER NAME
and 'the Israel of today,' whereas "old Atlanta" and "new Israel"
may imply chiefly a geographical dimension: 'the old part of At-
lanta' and 'the occupied territories of Israel.' Which adjectival
modifiers of place names resist the and which permit it with what
distinctions of meaning are questions still to be answered, and
whose answers will require that a large corpus of data be surveyed.
Similar problems regarding restrictive modifiers, especially ad-
jectives, arise with other kinds of proper nouns. For example, holi-
day names, for which the seems to be optional when a restrictive
adjective is present: "Hanukkah is (the) Jewish Christmas, and
Easter is (the) Christian Passover." In general, the use of the with
restrictive adjectives remains to be fully explained.
Proper nouns can be characterized by their cooccurrence with
the definite article, although the rule must be qualified, as has been
pointed out, when restrictive phrases or adjectives are present. A
number of other syntactic peculiarities have also been suggested
for proper nouns, none of which are reliable as definitions. It has
been observed, for example, that proper nouns do not freely accept
adjective modifiers (Long 1961:39; Strang 1968:114). While it is
undoubtedly true that adjectives occur more often with common
nouns than they do with proper nouns, such a statistical generaliza-
tion seems to be a consequence of the way proper nouns are used,
a matter of pragmatics, rather than an insight into the syntactic
nature of the word class. Adjectives do occur with proper nouns.
Although a relative clause, as in "They will nominate Teddy, who
is charismatic" is more to be expected than an attributive adjective,
as in "They will nominate charismatic Teddy," the latter is never-
theless grammatical. Moreover, familiar expressions like "wily Ulys-
ses" and "merry old England" are normal enough, and new epithetic
combinations are readily accepted, such as "spunky George Wal-
lace," which became commonplace in the news media shortly after
the assassination attempt. If there are limitations on the use of ad-
jectives with proper nouns, they seem to be matters of frequency
or to depend on the lexical items involved-and thus to be of only
limited interest to a grammarian.
Jespersen (1909-49:7.458) has proposed that proper nouns are
never used as predicate complements, a view shared by Lyons
(1966:213). In this idea Jespersen was consistently applying two
other proposals of his, namely that subjects are always more specific
than their predicate complements and that proper nouns are the
most specific of all words, from the conjunction of which it follows
that proper nouns cannot be predicate complements. The conclu-
sion can be maintained in the face of sentences like "the author of
this book was Otto Jespersen" only by the device of saying that in
such cases the proper noun is really the subject and the apparent
subject is really the predicate complement (Jespersen 1924:153).
The claim is thus vacuous, for counterexamples are simply reinter-
preted in conformity with the original claim. Jespersen's view of
proper nouns as the most specialized of all words, coupled with his
opinion that modifiers must be less specialized than the words they
modify, also led him to the conclusion that proper nouns are never
modifiers (Jespersen 1924:77). This conclusion is equally vacuous,
since Jespersen would account for exceptions like a Pinter play by
maintaining that in this use Pinter is not a proper noun.
Archibald A. Hill (1958:232), on the contrary, recognizes that a
proper noun can be used to modify a noun in immediately pre-
nominal position, as in "the good old Smith house," a position
shared by common nouns: "the good old stone house." Hill says that
common nouns can also occur as modifiers in predicate position:
"The good old house is stone"; whereas proper nouns cannot: *"The
good old house is Smith," in which construction a genitive is instead
required: "The good old house is Smith's." This distribution is, as
Utley (1963:172-73) notes, a genuine characteristic of adjunct use;
it is not, however, a unique characteristic of the adjunct use of
proper nouns. When the proper noun has a different sort of re-
ferent, for example "the good old Miami house," neither paraphrase
is possible: *"The good old house is Miami" nor *"The good old
house is Miami's." Instead, what is required is "The good old house
is in Miami." Similarly, for "the beach house," the acceptable para-
phrase is "the house is on the beach," the difference in preposition
being attributable to the complex, but for the present purpose irrel-
evant, system of locative expressions in English. So too, for "the
blacksmith shop," the paraphrase is "The shop is the blacksmith's."
The form of the predicate expression is determined not by whether
the noun modifier is proper or common but by whether its referent
is a person like Smith or blacksmith, a place like Miami or beach,
or a material like stone. The attributive position of such modifiers
disguises differences that appear in the predicate position. If English
were more Germanish, it might be possible to talk about *"the good
old in-Miami house," thus preserving the distinction.
ON DEFINING THE PROPER NAME
Dwight Bolinger (1971:61-66) has pointed out a way in which
proper nouns are like unmodified definite common nouns. He ob-
serves that certain kinds of adverbial modifiers can come between
the verb and its object only if the object is a modified noun phrase:
"They took with them the youngest boys" is acceptable, but *"They
took with them the boys" is not, or at least is less so. The same dis-
tinction is found with proper nouns: *"They took with them John"
is unacceptable, whereas "They took with them the youngest John"
is permissible. To the extent that proper nouns share such restric-
tions with common nouns, they are seen to be essentially the same
sorts of words, with only the special restriction on the occurrence
of the definite article to distinguish them.
Some recent generative grammars have distinguished proper from
common nouns by introducing a binary feature such as COMMON
(Jacobs and Rosenbaum 1968:60; Stockwell et al. 1968:982),
PROPER (McCawley 1968:134), or UNIQUE (Chafe 1970:112) in the
lexical entry of every noun, so that boy might be marked + coM-
MON, whereas Bill would be marked COMMON. The danger in such
a purely taxonomic feature is that it may be mistaken for an ex-
planation. It is what Lakoff (1971:ii) has called fudge: "A fudge,
as any student of the physical sciences can tell you, is a factor that
you add to what you've got to give you what you want to get. . .
One of the principal ways of fudging is to give a phenomenon a
name rather than to provide an adequate description of it." A fea-
ture like PROPER explains nothing. At best it is a flag marking a part
of the grammar that needs attention; at worst it disguises the need
for description. It is still necessary to discover whether each word
is common or proper, to specify the criteria by which the discovery
is made, and to decide what effect the difference has. Generative
grammar has so far provided no such insight into the nature of
It has been proposed that generic nouns are in some respects like
proper names (Christophersen 1939:59; Hill 1966). While it is pos-
sible to suggest somewhat vaguely that they are "proper names for
a whole class," there is no significant formal parallelism. The generic
use of a noun, as the chart below shows, can be indefinite singular
or plural, definite singular, or unspecified singular. What may
be thought of as the typical case is a unit noun like horse.
Man is an exceptional unit noun, in that it lacks the definite
generic ("The man is a biped" can only refer to some specific man),
INDEFINITE SINGULAR INDEFINITE PLURAL
A horse is a quadruped. Horses are quadrupeds.
A man is a biped. Men are bipeds.
A Datsun is a compact. Datsuns are compacts.
Milk is a liquid.
A Sue is a female. Sues are females.
DEFINITE SINGULAR UNSPECIFIED SINGULAR
The horse is a quadruped.
Man is a biped.
The Datsun is a compact. Datsun is a compact.
having instead a form that is unspecified for the category of definite-
ness. Datsun and other names for products have the full range of
possibilities. Mass nouns like milk have only a single generic form:
they lack number contrast in any use and are without a definite
generic form. Proper nouns like Sue also have only indefinite gener-
ics, a peculiarity they share with some common nouns for which a
definite generic would be odd-plural nouns like trousers and ab-
stract nouns like thought and estimation.
It is perhaps worth noting, as already observed by Long (1961:
236), that citation forms are syntactically proper nouns. "Horse
has five letters" is syntactically like "Harry has five senses," and
"The accommodations on this page are all misspelled" is like "The
Algernons in this room are all misguided." The syntactic parallelism
is, of course, not matched by the semantics of the expressions, al-
though it may suggest that citations are something like "proper
names" for the utterance type of which they are citations.
It is also noteworthy that the proper/common distinction is neu-
tralized when a noun is used in the vocative function, as in "Listen,
Aristotle" versus "Listen, man." Any noun in vocative use is neces-
sarily definite; the person addressed may be assumed to know his
own name and to know which appellatives are applicable to him.
ON DEFINING THE PROPER NAME
Because indefinite reference is not a possibility under these circum-
stances, the formal contrast between it and definiteness is not made
at all. The result is that all vocatives appear to be proper nouns;
but since there is no contrast between the two word classes in this
function, it is preferable to say that the distinction is neutralized.
It is often assumed that the distinction between unit and mass is
not relevant to the class of proper nouns, usually because the latter
are thought to represent a prior subcategorization (all nouns are
either proper or common; all common nouns are either unit or
mass). Most proper nouns are indeed unit nouns, but there is at
least one group of words in which the categories of mass and proper
seem to intersect. Mass nouns are those that (1) have no plural,
(2) have a simple indefinite form without a, and (3) are modified
by words like much, little, less, and a lot of rather than many, few,
fewer, each, and every; thus: fun, much fun, not *a fun, *many
Names of languages have the foregoing characteristics. They re-
sist plural formation: *Frenches; reject the indefinite article: *a
Spanish; and accept mass modifiers: "much English, little Latin,
less Greek, and a lot of Arabic." Language names can depart from
these characteristics, for example in referring to Great Russian and
White Russian, one might speak of "the two Russians," and it is
possible to distinguish between "a prestigious Italian" and "an
Italian of the streets." But all mass nouns can be used in those ways,
as "a malicious fun" and "a fun that is innocent" are "two different
funs." There is no reason to deny that language names are mass
nouns. On the other hand, they have also the principal character-
istics of proper nouns. They require the definite article when they
are modified by a restrictive clause: "the Hebrew that I know," but
not when they are unrestricted: "I know Hebrew."
Thus it appears that the proper/common system and the unit/
mass system are overlapping categories for which nouns must be
independently specified. Although proper mass nouns are limited
in range, they do exist. Long (1969:112) noted that language names
belong to the class of mass nouns, but he drew from that fact the
conclusion that therefore they are not true proper nouns. Long's
conclusion seems to rest on his initial, traditional but unwarranted,
assumption that mass and proper are mutually exclusive categories.
Names of illnesses such as tuberculosis, appendicitis, influenza,
arthritis, and indigestion are possibly another group of proper mass
nouns. They are clearly mass nouns: all resist the plural and the
indefinite article. Because of the nature of the illnesses, degree
modifiers are semantically odd with some of them in some uses
(witness the strangeness of saying that someone has "a lot of appen-
dicitis" although it is not strange to say that there is "a lot of appen-
dicitis among the students" and it is normal to say that someone
has "a lot of indigestion"). It might be doubted that such words
should be called proper, but they generally reject the definite article
except when they are restrictively modified: "He has influenza and
arthritis," whereas other disease names do not: "He has the flu and
the gout." Still other disease names are not only common, but also
unit nouns: "a cold," "a plague." The fact that words belong to the
same semantic set is no guarantee that they will belong to the same
grammatical classes. The use of articles with disease terms is too
complex for an easy generalization about their grammatical class.
Although the syntax of names has been most thoroughly studied,
their morphology has also been given some attention (Long 1969),
particularly the morphology of special kinds of names, such as those
for store-front churches (Dillard 1968) or place names (McMillan
1949; Zinkin 1969). It is impossible, however, to say anything posi-
tive about the internal grammar of names that is true of all, for
different kinds of names have altogether different internal structures.
The analysis of English personal names into given name, first name,
middle name, family name, maiden name, married name, genera-
tional tag, and so forth-or of Icelandic names into given name and
patronym or of Roman names into praenomen, nomen, and cog-
nomen-has no application to names of places or of other things.
It is not even the case that names have to be morphological nouns
or noun phrases, an assumption sometimes explicitly (S0rensen
1958:156) or implicitly made, but easily falsifiable. Hockett (1958:
311) cites Menomini personal names like [ana-ma-nahkwat] 'under
a cloud' and [awa'nohape-w] 'he sits in a fog' as instances of proper
names that have internal structures alien to the noun class. Levi-
Strauss (1966:177) points out that one of the standard name pat-
terns among the Iroquois consists of a verb with an incorporated
noun: He-raises-the-sky, He-announces-victory, She-works-in-the-
house. It is not necessary to have recourse to the exotica of Amer-
indian tongues for such examples. The names of "convenience"
stores in any American town, for example U-Pak-Em, will show
nonnominal internal grammar, as do the U-Haul and U-Drive-It
ON DEFINING THE PROPER NAME
companies. The Dare to Be Great organization was in the news
recently under charges of fraud, presumably for daring too much,
and the I Can't Believe It's a Girdle girdle has been widely touted
on television. The names of racehorses are an unfailing source of
native exotica: More, Sincere, Unintentionally, Communicate, In
Focus, Little But Fast, Who's Afraid, Hurry Up Dear, and Gee
Judge have been reported by Long (1969:109, 120); racing forms
will furnish a good many others. Although such names have regular
nominal syntax, their morphology is exceedingly diverse.
Personal names are sometimes nonnominal in their etymology
and also in their synchronic grammar as far as they are subject to
internal analysis. Thus prepositional phrases underlie Attebury 'at
the borough,' Atwood, and Atwell; and verbal or adjectival expres-
sions are behind Golightly, Dolittle, Fullalove, Lovejoy, Makepeace,
Turnbull (Matthews 1966). The latter group of names reminds us
that not only names but appellatives also are not necessarily nouns
in their internal grammar, witness scarecrow, pickpocket, and spoil-
sport, whose simplest derivation traces them to verb phrases. It is
easy to exaggerate the oddity of exotic languages. During the six-
teenth and seventeenth centuries, the Puritan population of England
produced some fruity specimens (Bowman 1931:91), given names
like The Lord Is Near, From Above, or the justly famous Praise-God
Barebones and his even more memorably named brother If-Christ-
Had-Not-Died-For-You-You-Had-Been-Damned Barebones, known
to his familiars as Damned Barebones. Other worthies of the period
were Faint-Not Dighurst, Be-Steadfast Elyarde, and the doubtless
sincerely named Flie-Fornication, the bastard son of Catrus An-
drews. Accepted Frewen, archbishop of York, and his brother
Thankful are evidence that even the establishment shared the
vogue for giving characterizing names not limited to nominal
It has been repeatedly demonstrated in this chapter that the
grammatical distinction between proper and common nouns is not
isomorphic with the semantic distinction between names and non-
names. Words that belong to the same semantic set are treated syn-
tactically in different ways. The Great Salt Lake, the Everglades,
the Bastille, the Cid, and the Bronx are grammatically common
nouns, whereas Lake Erie, Yosemite, Alcatraz, Charlemagne, and
Brooklyn are grammatically proper. Yet all are names.
On the other hand, baseball terms like first, second, and third
base illustrate a discrepancy of a different sort in that semantically
they are not proper names but ordinary appellatives like outfield,
dugout, and mound. But whereas the latter are common nouns, re-
quiring an article, as in "He went from the outfield to the dugout
and then to the mound," the former are proper nouns used, when
singular and unrestricted, without the definite article: "He is on first
It is also obvious that grammatically defined proper nouns are
language-specific. Some other languages have distinctions like that
of English. Thus Hockett (1958:311-12) reports that in Fijian a
particle ko marks proper nouns in contrast to the particle na, used
with common nouns, so that na vanua levu means 'the big island,'
whereas ko vanua levu means 'Big Island.' Other languages, how-
ever, such as Latin, in which via longa can be either 'the long road'
or 'Long Road,' have no particles by which to distinguish proper
from common nouns and thus may not have the distinction as a
syntactic category at all.
Of those items that are generally recognized as proper names,
some have the internal structure typical of nouns, others do not;
some have an incompatibility with the definite article that marks
proper nouns, others are common nouns. These facts make it im-
possible to treat names as a simple grammatical class. It is neces-
sary to distinguish between names defined morphosyntactically-
proper nouns-and names defined in some other way (Sledd 1959:
244; Long 1961:503, 1969:107). The question that remains to be
answered is in what other way names can be defined.
SHE EARLIEST and far and away the most often used
definition of proper names describes them as words
used to refer to single, particular, or unique individuals-names
for classes that have only one member. This definition, which seems
to have originated among the Stoics, appears in the oldest of Western
grammars, that of Dionysius Thrax, and was popularized in
Europe by the Ars Minor of Donatus. Thus Dionysius says that a
proper noun is one which signifies an individual or particular sub-
stance (idian ousian), whereas a common noun signifies general
substance (koinen ousian); and Donatus's catechistical grammar
runs thus: "Qualitas nominum in quo est? Bipertita est: aut enim
unius nomen est et proprium dicitur, aut multorum et appellati-
vum." Here we have the two dominant themes of this definition: a
proper name applies to a single individual (unius nomen est) that
is a particular, unique being (idia ousia). The themes can be seen
again in the definitions of the seventeenth- and eighteenth-century
writers of universal grammars, such as the Port Royal grammar
(Lancelot 1660:29), which distinguishes between proper names as
expressions for single things and appellatives as expressions for sev-
eral similar things, or James Harris's Hermes (1751:345), which
holds that proper names are expressions for particulars as opposed
to general ideas.
More recently, a similar approach has been taken by many gram-
marians, of whom the following are representative: Jespersen
(1922:438) speaks of "proper names of the good old kind, borne by
and denoting only one single individual," thus suggesting that we
must deal also with a bad new kind of name that is not so re-
stricted. Bloomfield (1933:205) defines the class meaning of proper
names as "species of object containing only one specimen [Bloom-
field's italics]." Collinson (1937:21) believes that "a proper name
refers to a particular individual and to no other." So also, "A proper
name is a symbol directly identifying a single individual" (Chris-
tophersen 1939:62), and "A proper name, in so far as it remains a
real proper name, is a word which refers only to one individual
thing, usually a person or place" (Gardiner 1951:41). Gardiner
later retracted this definition (1954:22), but it continued to be held
by Coseriu (1967:279): "El nombre proprio . es siempre nombre
de un 'singular,"'" and by Palmer and Blandford (1969:45). Em-
phasis has been placed on the referential uniqueness of proper
names by Brown (1851:239), Curme (1935:1), Long (1969:109),
Chafe (1970:112), Stageberg (1971:283), and Quirk et al. (1972:
160), to mention some grammatical treatments of different kinds
A definition honored by such long and persistent use and exem-
plifying such uncommon accord between the ancients and the
moderns should carry great weight. Yet quite apart from the philo-
sophical problem posed by the notion of a unique individual, which
was acknowledged by Jespersen (1924:69), the definition does not
define the class of items it is intended for. Although all languages
certainly have means of referring to particular, single individuals,
it is not clear that, of those words we might want to call proper
names in natural human language, any uniquely refer to particular,
single individuals. On the contrary most of them clearly do not
Words like Andes and Antilles that are plural in form are no
particular problem. They have no singular *Ande and *Antille and
therefore may be taken as applying globally to that which they
designate (Jespersen 1924:64). They are not essentially different
in this respect from collectives like Mafia, which though formally
singular can take either singular or plural concord; the grammatical
number of a noun is largely an arbitrary way of interpreting the
number of things. Gardiner (1954:24) believed that words like the
Medes and the Angles were a more serious and indeed an insuper-
able problem, because they have singulars, a Mede, an Angle. If
such words are proper names, Gardiner was correct; they seem,
however, to belong with Ford and Volkswagen as appellatives, ex-
cept when used as names of the tribe, not tribal members.
Nor is it of any linguistic consequence that there are proper
names to which no actual persons or things correspond, names like
Xanadu, Pegasus, Mary Poppins, and Smaug. Such words pose a
ON DEFINING THE PROPER NAME
philosophical problem that has been resolved in various ways, in-
cluding the paradox of Russell's view (1905:479) that "a phrase
may be denoting, and yet not denote anything," or his other view
(1940:37) that if the thing a "true proper name" purports to name
does not occur, the name is meaningless. For linguistics, however,
the problem does not exist, for the grammarian is not concerned
with the ontology of linguistic referents. For him it is enough to
know whether a speaker intends that two terms shall have the same
or different referents; whether those referents are of this nature or
of that is a problem that the grammarian is happy to leave to the
philosopher. With the new-found respectability of mentalism in
linguistics, some linguists no longer share Mill's conviction (1843:
1.2.1) that "names are names of things, not of our ideas," and talk
rather about "items in the speaker's mental picture of the universe"
in distinction to "real things in the universe" (McCawley 1968:138),
but that may be more a convenience than an ontological commit-
ment. For the linguist, Xanadu has as good a referent as Canada,
and in neither case does it make any difference to him what sort
of referent it is.
A more serious problem is that most personal names and many
place names are in fact applied to any number of different persons
and places (Strawson 1950:188; Long 1961:229). There is many a
Tom, Dick, or Harry in this world. There are more Athenses than
the one in Attica, and more Cairos than the one in Egypt. As
Leonard Linsky (1963:76) has paronomastically observed, "Proper
names are usually (rather) common names." It is possible, to be
sure, to point to an occasional proper name that seems to have a
unique referent-Popocatepetl is one-but there is nothing to stop
Congress tomorrow from deciding to rename Pike's Peak Popocate-
petl, thus destroying the uniqueness of reference. When a proper
name refers to a single individual, it seems to be an accident of
use, a fact of parole, and not a matter of the language system.
There is nothing in English or apparently in any other natural
language that requires a name to have unique reference.
In view of the obvious fact that most personal names and many
place names have more than one referent associated with them,
the position that proper names have unique reference can be main-
tained only by recourse to the notions of homonymy or polysemy.
That is, we can say either that John as applied to John Adams and
John as applied to John Baker are different words that happen to
be pronounced alike or that they are the same word with different
meanings. Two items with the same pronunciation but different
meanings may be said to be the same polysemous lexical item if
they have the same grammatical distribution, that is, if they belong
to the same part of speech and follow the same rules of syntactic
concord. Otherwise, if the meanings are correlated with some
grammatical distinction, the items must be accounted different, al-
though homonymous, words. For example, iris belongs to the same
part of speech and in other respects is grammatically unchanged
regardless of whether it is taken in the sense 'flowering plant' or the
sense 'part of the eye.' It is therefore polysemous--one word with
several senses. Pupil 'student' and pupil 'part of the eye,' on the
other hand, are two homonymous words because the former requires
personal and animate gender concord (who, he or she) whereas the
latter requires impersonal and neuter concord (which, it). Ety-
mology, the touchstone most often used by dictionaries in deciding
whether two items are the same or different words, is irrelevant to
the synchronic question.
The distinction between polysemy and homonymy is, however, a
Tweedledee-Tweedledum one for the problem of John as applied to
John Adams and John Baker. Both approaches maintain that a
vocable like John has a different sense as it is applied to different
referents. This view in one form or the other has been suggested
by such contemporaries as Gardiner (1954:16), Hockett (1958:
312), S0rensen (1963:83-86), and Chafe (1970:112, 274).
The view, however, has some consequences that are unacceptable
either theoretically or practically. If we say that homonymy is in-
volved, we will be forced to conclude that there are as many differ-
ent words John as there are or have been or will be persons bearing
the name. As the person John" differs from the person John7, we
shall have to say that the name John applied to the first is a differ-
ent word from the name John applied to the second. This position
makes the vocabulary infinite and thus in principle indescribable.
Genuine cases of homonymy, like that of pupil, involve two or three
or some relatively small number of distinct words that are pro-
nounced alike. They do not involve an infinite number of different
words. Polysemy is no better as an explanation, for by it we must
say that the word John has as many meanings as there are, were,
or will be persons so named. The vocabulary is still unlimited and
indescribable because every proper name has a potentially infinite
ON DEFINING THE PROPER NAME
number of meanings. It is as though we should say that iris has not
just a few senses, including the two mentioned above, but as many
senses as there are individual flowers of the family Iridaceae. The
consequence would be to make definition impossible. In either of
these approaches, meaning is confused with reference, since there
is assumed to be a one-to-one relationship between meanings and
referents. The distinction between that which a word means and
that to which it can be used to refer is an important one that has
been painstakingly made by students of language such as Frege
(1892) more recently and in the thirteenth century by Peter of
Spain (Mullally 1945:3-5). It should not be lightly abandoned, and
certainly not for the sake of maintaining an otherwise faulty defi-
nition. The notions of homonymy and polysemy, as they have been
applied in onomastics, provide no tools for handling the data of
proper names; on the contrary, they make description impossible.
Instead of explaining the use of proper names, they avoid explana-
tion by confusing meaning and reference-a confusion that would
not be permitted for appellatives.
It would appear that the traditional view of proper names as
words for unique individuals cannot be maintained, or can be main-
tained only at a very high cost-that of increasing the size of the
vocabulary beyond all manageability. But if that is right, we might
wonder how it is that the traditional view has been held so long
and by so many. There are two main reasons. First, what tradition-
alists said was in many cases not what they meant. The plain sense
of the traditional definition is that a proper name is a term for one
and only one particular referent. What seems rather to have been
meant is that within any given discourse, the speaker may use a
name with the intention of identifying a particular individual and
that he further assumes the person spoken to will succeed in identi-
fying that individual from the name. This is quite a different thing
from saying that each name of itself has unique reference, inde-
pendent of its verbal and situational context. What those who have
spoken of a unique referent seem often to have intended is the
grammatical category of definiteness.
It has been proposed that proper names are inherently definite
(Jespersen 1924:65; Bloomfield 1933:205; S0rensen 1963:92; Postal
1966:204; Chafe 1970:198), just as personal pronouns, demonstra-
tives, and adverbs like here, there, now, and then are inherently
definite (Leech 1969:277). But this proposal rests on the unstated
assumption that words are proper names only if they have definite
reference, and is thus really a definition in disguise. If proper names
are defined in some independent way and then it proves to be the
case that they are always definite, an insight into language structure
has been gained. That, however, is not the case. When we consider
paradigmatic cases of the proper name-words like George, Jones,
Moscow, and Siam-it is clear that they are usually, but not in-
variably, definite (Langendoen 1970:124), because of expressions
like "a George" and "any Moscow."
It might be said that for many names, definite reference is the
unmarked state, whereas for most appellatives, indefinite reference
is the unmarked state. Such a conclusion is supported by the formal
syntactic signals of English grammar. Thus when appellatives occur
without determiners, for example the mass noun water in "He drinks
water" or the plural unit noun men in "They employ men," the
reference is indefinite, whereas a determiner-less name like Bill in
"They employ Bill" is definite in its reference. This formal distinc-
tion in modern English parallels what seems to be a widespread
feeling about the difference between proper names and appellatives
even in those languages like Greek or Latin which lack such formal
signals. If this markedness relation is correct, it explains both why
grammarians since the Stoics have thought that proper names de-
noted unique individuals (that is, had definite reference) and why
the formal signals of modern English should have developed as
they did. It can perhaps be accepted as a general principle of his-
torical change that a language tends rather to develop overt mark-
ings that reflect covert categories than to lose them or to develop
overt markings at variance with covert categories. A characteristic
distinction between proper names and appellatives is that we expect
the former to be definite and the latter indefinite rather than the
other way around. Bill, water, and men are simpler not only in
form but also in meaning than a Bill, the water, and the men.
That names are not invariably definite can, however, be seen in
a language like Armenian, in which a name such as Arto may occur
either with or without the definite article (Der-Houssikian 1970:
8-9). Thus in the sentence Arton egav 'Arto came' the name re-
quires the suffixed definite article -n in normal circumstances, in
which the speaker assumes that the addressee knows the identity
of the referent. If the situation is such that the addressee cannot be
assumed to know the identity of the referent, for example in "I
ON DEFINING THE PROPER NAME
have a friend called Arto," the article is not used. Expressions of
the type "named, called X" that include a citation of the name are
linguistically special in some ways to be described in chapter 6,
but it is worth noting here that also in English such expressions
are basically indefinite, alternating, as they do, with an overtly
signaled indefinite form of the name: "Someone called George is
here to see you" and "A George is here to see you" are equivalent
expressions with respect to indefiniteness.
It is not the case, however, that all proper names use an unmarked
form for definite reference. The Alamo, the Appalachian Mountains,
and the Lake of Geneva are proper names whose definiteness has
to be marked. In an attempt to save definiteness as a defining cri-
terion for proper names, Waismann (1965:196) has proposed that
names are single words rather than "syntactically formed groups of
words," thus excluding an expression like the murderer of Caesar
as a proper name while allowing Brutus. This distinction would,
however, exclude also the Lake of Geneva, as clearly a group of
words with a syntactic structure, the Appalachian Mountains, Lake
Leman, Mount Pisgah, the Alamo, and quite likely Marcus Junius
Brutus as well. Such paradoxes can be avoided by recognizing units
on more than one linguistic level. Semantically, all of the items
cited in the last sentence are proper names. Those that take the
are common nouns or noun phrases; those that do not are proper
nouns or noun phrases. Definiteness is then a category shared by
names and appellatives. Many names are like Lake Leman in being
morphologically unmarked for definiteness; others are like the Lake
of Geneva in being marked for the category. The categories PROPER
and DEFINITE are not equivalent. Reinterpreting the traditional view
of the uniqueness of proper names as a matter of definite reference
will not account adequately for the diversity of use of such names
in ordinary language.
The second reason that the traditional view has held sway is that
it has confused ordinary language with the specialized register
used in logical discourse. As already noted, the earliest grammatical
speculations of the Hellenic world did not distinguish purely
linguistic description from philosophical considerations. Grammar-
ians are still to some extent influenced by the categories of logic,
especially in dealing with names.
The logician finds it necessary to have some way of identifying
what are, within his universe of discourse, unique referents. He
consequently specifies such ways and often calls what is so defined
a proper name, although such "proper names" coincide very im-
perfectly with the class of items that the grammarian might want to
designate. Frege (1892:57), for example, defined proper name as a
sign for a definite object, thus embracing both Brutus and the
murderer of Caesar within the class. Russell has said a great many,
not always consistent, things about names; but for him this is as
good a proper name as Homer, indeed rather a better one (Russell
1919:179; 1940:116-37), and a name is an indivisible symbol
"whose meaning is something that can only occur as subject, i.e.
... an 'individual' or a 'particular.' The latter view has been taken
up by a number of grammarians (Jespersen 1909-49:7.458; 1924:
153; Lyons 1966:213). Although it may be true by definition for
"logical" languages, it is false for ordinary language on two counts:
proper names do not always name particulars since they occur
grammatically indefinite, and particular terms are not limited to the
subject function of ordinary linguistic structures.
Russell (1905; 1919:179) and Quine (1950:220-24) developed a
method of eliminating singular terms, that is, logically proper
names, from logical discourse by the expedient of making predi-
cates out of them. Thus Homer may be expressed something like
"There is a y such that for every x, x is-Homer (Homerizes) if and
only if x is y." A somewhat similar use of variables has been pro-
posed for natural languages (Bach 1968:110-11). Although the
idea that the deep structure of human speech is the predicate cal-
culus may seem not far removed from Householder's (1971:xi)
tongue-in-cheek proposal that the fundamental form of a language
is Morse code, more will be said about it in chapter 6.
Linguists have often benefitted from the work of philosophers,
who see the subject matter with which both are concerned from
a different and potentially fresh point of view; however, it is not
clear that the philosophical concern with particulars has any rele-
vance to natural language. The linguist Charles Hockett (1958:
312) has cogently remarked that "to logicians, a 'proper name' or
'proper noun' is a symbol which designates an entity of which there
is only one. .. In actual languages there are no forms which can
be so described, save possibly through pure accident." And a
logician concerned with ordinary language, Leonard Linsky (1963:
76), has also recognized that proper names need not involve a
unique relationship between the name and the one named: "What
ON DEFINING THE PROPER NAME
is indeed necessary, if I am to make a definite assertion, is not that
one person only be named 'Tommy Jones,' but that I be referring
to just one person, however many others there may be with the
same name as his." For the description of natural languages, as op-
posed to purely logical systems, we must add that for definiteness,
it is necessary not only that the speaker be referring to just one
person, but also that he have reason to believe that, within the con-
text of the language act, the person addressed will be able to iden-
tify the one intended. If the speaker is referring to a specific person
but has no reason to think the addressee can make a correct identi-
fication from the name, he refers not to "Tommy Jones" but rather
to "a Tommy Jones" or the like. Definiteness reflects the speaker's
assumptions about the knowledge of those he speaks to and about
the situation in which the name is used.
In a contrast of proper names with common appellatives, there is
often a confusion between the linguistic system and the use of that
system. Thus Gilbert Ryle (1957:139) believes "it is obvious that
the vast majority of words are unlike the words 'Fido' and 'London'
in ... that they are general. 'Fido' stands for a particular dog, but
the noun 'dog' covers this dog Fido, and all other dogs past, present,
and future, dogs in novels, dogs in dog breeders' plans for the
future, and so on indefinitely." What Ryle believes to be obvious is
instead a muddle of two different aspects of language. Dog covers
all the dogs Ryle indicates-living and dead, actual and projected,
real and fictional-as an item in the lexicon, a linguistic abstraction.
Fido stands for a particular dog only in some particular context,
as when a woman says to her husband, "You'd better take Fido for
his walk." In such a context, however, dog is just as particular as
Fido. Thus the woman might have said instead, "You'd better take
the dog for his walk," without intending all those dogs whose spec-
ters Ryle invokes. Conversely, Fido, considered abstractly, as a
characteristic name for a dog, is a general term too. Admittedly it
is not so general as dog, but that is because Fidos are a proper sub-
set of dogs, just like terriers or one-eyed, tailless hunting dogs.
The use of a word like dog or Fido in some particular context
should not be confused with the word as an abstraction, a lexical
unit available for use when it is needed. To compare dog as an
isolated lexical unit with Fido in a particular contextualized use is
to compare incommensurables and reveals nothing about the dif-
ferences between the two words.
The contrast drawn here between the linguistic system consid-
ered in the abstract and a particular application of the system re-
flects in part the Saussurian distinction between langue and parole
and in part the logicians' distinction between the mention of a form
and its use. Gardiner (1954:8-10) draws the same contrast when he
distinguishes between "disembodied" and "embodied" proper names,
which are, respectively, names considered in isolation and names
taken in some context. It is likewise reflected in the familiar dis-
tinction between MEANING and REFERRING, used by philosophers
and grammarians alike since Peter of Spain (Pope John XXI) drew
the distinction and called it SIGNIFICATION versus SUPPOSITION. A
linguistic form has reference to some thing or state of affairs only
when it is used in a particular context, whereas it has meaning even
in isolation, merely by virtue of its being a linguistic form. If a
word's meaning is the set of conditions under which it can be used
appropriately, its reference is what, in some particular use, it is
being used for.
Two tokens (Fido as uttered by one person and Fido as uttered
by another or by the same person a second time) count as the same
use of the type (the lexical form Fido) provided they have the
same reference, and as different uses otherwise. There is thus a
three-way distinction between a type (which is a matter of the
language system), the use of a type (which is a way the system
can be applied), and the utterance of a type (a token, a speech
act). As Strawson (1950:171) observes, we cannot say the same
things about these three. Those who contrast the "uniqueness"
of proper names with the nonuniqueness of common appellatives
are generally contrasting a use of the former with the latter as a
type. Nothing whatever is thereby shown about the forms, for
uniqueness is a characteristic not of the meaning of types, but of
a particular referential use. In this respect names do not differ
from appellatives, which also have unique uses, as in "this dog,"
"my father," or "a certain book," provided these phrases are con-
sidered in contexts in which they have genuine use and are not
merely mentioned, as in this sentence.
"Disembodied" proper names considered in isolation are, as
Gardiner observed (1954:8), the subject of linguistic studies con-
cerning their etymology, their frequency of occurrence, and other
intralanguage facts. "Embodied" proper names, as they are used in
context, are treated in encyclopedias, biographical dictionaries,
52 ON DEFINING THE PROPER NAME
gazetteers, and so forth, where the subject of interest is not the
word itself, but the things named by it. The distinction between
lexical and encyclopedic facts, which has been dealt with by Gove
(1965), is not always easy to make, but it is of both theoretical and
practical importance, especially to onomastics, for any effort to de-
fine a proper name by its referents is to move out of language into
other areas of study.
Language has reference, and some linguistic expressions (singu-
lar terms) are used to refer to single unique individuals: George
Washington Carver, my eldest child, the tallest living woman, the
car, he, whereas other expressions (general terms) do not purport
to be so specific in their reference: most scientists, a girl, any house,
someone. This difference in reference is, however, clearly not
coterminous with the distinction between names and appellatives.
Semantic Names: The Degree of Meaning
LOSELY associated with the view that proper names
have singular reference is the view popularized by
John Stuart Mill that names are words without signification, mean-
ingless marks by which one thing is distinguished from another.
Thus, boy has a meaning, something like 'nonadult male human
being,' and refers to various creatures, whereas Bill has no mean-
ing, but merely refers to some creature. In Mill's terminology (1843:
1.2.5), a proper name is nonconnotative, that is, it identifies an
individual without imputing any attributes to him. According to
Mill, Cratylus was wrong, for Hermogenes could bear one name
as well as any other, all being equally meaningless.
Mill's view has been held also by Christophersen (1939:59),
who says that names lack any conceptual content and thus indicate
an object without implying a description of it; by Strawson (1950:
188), who holds that names have neither descriptive meaning like
common appellatives nor general rules like pronouns and demon-
stratives, but are governed by ad hoc conventions that have to be
established for each application of the name to a different person;
and by Togeby (1951:215), who would have names stand for in-
dividuals rather than semantic classes; and also by others who
phrase the matter in still other ways but whose views are essentially
those of Mill. It is customary to speak of this definition of the proper
name as Mill's, and certainly it is he who is chiefly responsible for
its modern use; however, its substance can be found in earlier
authors, such as John of Salisbury's "Nominantur singularia sed
universalia significantur" (cited by Jakobson 1972:78)-Individual
things are named, but general classes are signified.
Some who have shared Mill's view of names as nonsignificant have
rephrased the distinction between them and appellatives as one of
intensional versus extensional classes (Johnson 1921:94). A class
ON DEFINING THE PROPER NAME
that is defined by specifying some attribute or combination of attri-
butes shared by all and only the members of that class is intensional,
for example, a class consisting of all men who are over six and a
half feet in height, are bald, and have an even number of fingers on
their left hands. The membership of such a class may be unknown;
indeed it may have no members at all. But because the defining at-
tributes are clear, if anyone with those attributes exists
he will belong to the class. On the other hand, a class that is
defined simply by listing its members is extensional, for example,
a class consisting of Swami Vivekananda, strawberry jello, and
the first sentence of Jane Austen's Pride and Prejudice. The mem-
bership of this class is known, though it is doubtful that there
is any attribute in common to the three members. Intensional classes
are defined by specifying general characteristics, extensional classes
by listing membership. An appellative like boy would label an in-
tensional class, of which the defining attributes are 'human, male,
nonadult.' A proper name like Bill, however, would label an exten-
sional class, which has no defining characteristics-no connotative
meaning-but whose membership can only be listed: Bill Bailey,
Bill Buckley, Bill Hitchcock, and so on.
For Mill, proper and common names are alike in denoting things
but differ in that common names, in addition to denoting, imply
attributes. For others (Austin 1961:7; S0rensen 1963:14), the two
kinds of words differ in that, whereas proper names denote without
connoting, appellatives connote without denoting directly. This ex-
tension of Mill's view makes proper names a thing apart from other
nominal terms in language. It makes of them, in fact, something
very much like the variables x, y, z of mathematics and logic. Proper
names have been likened explicitly to such indexical symbols
(Hockett 1958:312; McCawley 1968:138); and Sampson (1970:
103-4) has proposed that names do not belong in the lexicon of a
language at all, but rather in a separate component, the "topicon,"
whose function is to account for referential identity and difference
within a discourse. Treating proper names as not part of the lexicon
avoids one problem that otherwise arises: if all names are equally
meaningless, then every name is presumably a synonym of every
other. This conclusion has been accepted by Togeby (1951) but
is surely a reduction ad absurdum. The fact that substituting one
name for another within a sentence can radically alter the meaning
of the sentence and its truth value suggests that names must have
SEMANTIC NAMES: DEGREE OF MEANING
nonsynonymous meanings. If names are not part of the lexicon, as
Sampson would have it, we cannot expect them to be subject to
categories like synonymy any more than x, y, or z.
There are other problems with Mill's view that cannot be so
easily disposed of. According to his definition, the application of
any name to an individual is purely arbitrary, being quite independ-
ent of any characteristic of the person, and thus one name is as
appropriate as another. For some linguists, the arbitrary imposition
of a name is the decisive criterion by which the class can be defined:
"Entscheidendes Kriterium eines Namens ist, dass er willkiirlich
gewahlt bzw. gegeben wird" (Witkowski 1964:50). In one sense,
most words, or more precisely morphemes, are arbitrary, in that they
are noniconic, the exception being onomatopoeic terms, although
even words like crash, moo, and hiccup are conventional since there
is no necessary relationship between their sounds and meanings.
The arbitrariness that is imputed to proper names is, however, of
a different sort. If a speaker of English acquires as a pet a domesti-
cated animal that chases mice and purrs, his decision to refer to
it as a cat is not arbitrary, but reflects his knowledge of the English
language; on the other hand, his decision to call it Miko may be
wholly arbitrary. Thus, the arbitrariness of proper names applies
to their initial bestowal on an object.
Yet there are several respects in which not all names are arbitrary.
Many speakers would consider it odd to name a cat Fido or Rover.
Gilbert Ryle (1957:137) agrees with Mill in supposing "from the
fact yonder dog is Fido, no other truth about him follows at all.
No information is provided for anything to follow from. Using a
proper name is not committing oneself to any further assertions
whatsoever." But Ryle's example, in fact, suggests a different con-
clusion from the one he draws. The fact that yonder entity is Fido
may not have any necessary logical implications, but the ordinary
use of the name Fido in ordinary language certainly presupposes
that the entity so named is a dog, and not a cat or a boy. Any other
use of the name would be odd. A few years ago, a song entitled "A
Boy Named Sue" was popularized by Johnny Cash; its lyrics de-
scribe the hard but educative life led by a boy whose father had
nothing to give him but a name that would ensure he would learn
to defend himself. If names were really arbitrary, the song would
have no point. McCawley's (1968:139) argument that "Gwendolyn
hurt himself" is odd in the same way as "My neighbor hurt him-
ON DEFINING THE PROPER NAME
self," when the latter is said of a woman, ignores the fact that the
first is odd in isolation, without reference to any context whatever,
and the second is not.
To be sure, not all given names are distinctive in gender. Some
are sexually ambiguous: Evelyn, Beverly, Tracy, Robin, and Dana,
for example (Prenner 1942). In the South, hypocoristic forms of
male names, such as Tommie, Jimmy, and Johnnie, are usual for
females, although often in parataxis with more conventionally
female names: Tommie-Sue, Johnnie-May. Imaginatively invented
names and the use of family names (often the mother's maiden
name) as first names for both males and females further blur the
distinction between the gender of given names. The distinction,
however, exists, as demonstrated by the fact that exceptions to it
are causes for comment. Names like Sue, and Fido as well, do have
meaning and imply something about the ones named.
Family names present a different kind of nonarbitrariness. To say
that a name like Miller or Smith is arbitrarily given (Christopher-
sen 1939:60; Witkowski 1964:49-50) is misleading, and an argument
to the effect that persons called Miller or Smith need not be millers
or smiths is merely a red herring. Whatever the provenience of the
name, no ordinary user of the language supposes that a man called
Miller grinds grain for a living, but ordinary users do suppose, with
considerable justification, that Mr. Miller's father was also called
Miller, and likewise his brothers, his unmarried sisters, his wife,
and his children. Family names are not, as a rule, arbitrarily given
to an individual, nor are they capriciously changed, although they
may be for due cause and with respect for appropriate procedures
(Smith 1969; Ashley 1971).
There have been occasions, even in recent history, when such
arbitrariness was characteristic of family names, for example among
the black population of the South in the immediately postbellum
years. Former slaves, in assimilating to the structure of white society,
arbitrarily took family names and often new first ones as well; more-
over, they changed those names at will, so that a freedman might
take on and put off a variety of names or even bear several simul-
taneously. "For instance, a boy entered school under the name of
Joseph Marshall; the boys called him Marshall Black; and the name
given him by his parents, and by which he was called at home,
was Joseph Black Thomas" (Donald 1952:151). Each member of
the family might have a different last name. If such an arrange-
SEMANTIC NAMES: DEGREE OF MEANING
ment were the normal one for "family" names, they could be said
to be arbitrary. That arrangement, however, was clearly not normal,
but rather a transitional one in which members of the population
that had formerly had no family name system were acquiring one;
in the process they treated "family" names like given names, which
need not be shared by members of a family unit, and both like nick-
names, which can be freely changed at the whim of the name-
Nicknames, although in one respect at the opposite pole from
family names in that they are neither customarily passed on to prog-
eny nor stable during an individual's lifetime, are also usually
nonarbitrary, albeit in a different way. Minnesota Fats was not so
nicknamed arbitrarily. If he had been called instead Slim, the
naming principle (lucus a non lucendo) would have been different
but not the nonarbitrariness of the name. Some bynames may be
arbitrary, but they are not the typical sort; indeed, the usual
function of the nickname is to suggest some characteristic of the
person named, that is, to be nonarbitrary.
There are many peoples among whom the formal given name can-
not be arbitrary; personal names, for example, may be required
to contain a reference to the clan's totem. Thus typical names for
members of the Black Bear clan of the Osage tribe are Flashing-
eyes, Tracks-on-the-prairie, Black-bear-woman, or Fat-on-the-skin;
names with allusions to another clan's totem would be impossible
(Levi-Strauss 1966:173-76). Also nonarbitrary are the day-names
mentioned in chapter 1 and ordinal names with the sense 'first born'
and the like (Levi-Strauss 1966:189-90). Even in American naming
practices, some families have a tradition of giving the eldest son
the name of his paternal grandfather; such "grandfather names"
may be bestowed with a regularity that is beyond arbitrariness.
An important and much studied sort of nonarbitrary proper name
is the characterizing name (also known as label name, attributive
name, or charactonym) used in literary works. Examples in the
English tradition are as old as Unferth (that is, 'Marpeace') of the
Beowulf epic and as recent as J. Alfred Prufrock, Frodo, or Alex-
ander Portnoy, with such intervening notables as Chaucer's Man
of Law's Constance, Falstaff, Lady Wishfort, Mrs. Malaprop, Pippa,
Arthur Dimmesdale, and numerous others. Such names have been
treated by Robinson (1972) and Green (1972), and in the 1968
issue of Names (vol. 16, no. 4) devoted to literary onomastics.
ON DEFINING THE PROPER NAME
To turn from personal to place names, it is sometimes suggested
that names like Mont Blanc, Dartmouth, and Land's End are impure
proper names because their parts are meaningful and appropriate,
and thus they are not wholly arbitrary (Gardiner 1954:42; Shwayder
1963:56). These words indeed exemplify another kind of non-
arbitrariness in names. The particular problem they raise is not
limited to onomastics, however, for it does not differ essentially from
the semantic problem posed by any idiom-an expression with
meaningful parts whose total meaning cannot be predicted from the
meanings of those parts-for example, nominal expressions like
bluebird, scarecrow, bull's eye, verbal combinations like look out,
come around, call on the carpet, or even whole clauses like A
rolling stone gathers no moss. All idioms, including proper names
of the kind mentioned, pose an interesting problem in the relation-
ship of the lexicon to the grammar, but it is not a specifically
The notion that, because proper names are arbitrary, expres-
sions like Mont Blanc, Sue, or Slim are names to a lesser degree or
not names at all is, as Utley (1963:151) observed, "a kind of
categorical obsession," completely at variance with ordinary lan-
guage. To take words like Vercingetorix or Popocatepetl as better
examples of proper names than Sue and Dartmouth because the
former are arbitrary, at least to most English speakers, whereas the
latter have some descriptive value, is to ignore the obvious fact that
most names have some motivation. Those that do not are the ex-
ception, not the paradigmatic case. Indeed, the notion that any
name is arbitrary and unmotivated is a relatively recent one. As
Elsdon Smith (1966:492) has pointed out, "In the past people
simply did not confer names from sheer whim and without any
meaning. It would be contrary to man's nature to name the objects
of his thoughts by sounds which conveyed no meaning to him or
others." There are, to be sure, names that spring from whim-the
pungent examples collected by Thomas Pyles (1947, 1959) are good
illustrations: Glathu, Aletrice, Juhree, Naul, Twyla, Zazzelle, Onan,
Phalla, and Merdine, which one hopes were unmotivated on the
part of the parental name-givers. Still, this kind of naming seems to
be an aberration. The fact that names are keys to the past for the
historian and the philologist is evidence that they are more than
meaningless index symbols.
It has also been maintained by Bertrand Russell (1940:116-37)
SEMANTIC NAMES: DEGREE OF MEANING
and Hockett (1958:312) that proper names are like substitutes,
such as personal or demonstrative pronouns, in serving a deictic
rather than a descriptive function, since they are supposed to
point out rather than characterize their referents. Although there
is some, albeit unclear, likeness between proper names and pro-
nouns, there are important differences. Russell himself has observed
that "the word 'men' is applicable to all the objects called severally
'a man,' but the word 'these' is not applicable to all the objects
severally called 'this' on different occasions." In this respect, proper
names are more like appellatives than like pronouns; for when plu-
ralized (Cairos, Georges, Joneses), they are applicable to all the
objects severally called by the singular name (Cairo, George, Jones).
Just as "The man is mortal" is implied by "All men are mortal," so
"George is intelligent" is implied by "All Georges are intelligent."
"This is good," on the other hand, is not implied by "These are
good," unless we have knowledge derived from some other source
that the "this" of the first sentence is one of the "these" of the second
sentence. Such independent knowledge about the referents is not
required for appellatives or for proper names.
In support of the proposition that names have no meaning,
it has been averred that they are not translated (Vendler 1967:38;
Manczak 1968a:206; Zabeeh 1968:69). But it is by no means clear
that the assertion is correct. On the contrary, names of different lan-
guages correspond in three ways that can be called translation.
First there is the treatment of names like Krasnaya Ploshchad,
the name of an architectural feature in Moscow, which is
rendered into English as a calque, Red Square. This is translation
of a name in the most obvious and least interesting sense. It is
essentially the same thing we do when we say that Mont Blanc
means 'white mountain,' Peter means 'a rock,' and Emanuel means
'God with us,' although in these cases the etymological calque is
not used as a name. In the case of Red Square or Hall of Supreme
Harmony (T'ai-ho-tien) or the Yellow River (Huang Ho), it is
hard to see how one can avoid saying that the name has been
translated. Indeed, the names of one class, the titles of books, songs,
paintings, and so forth, are regularly so translated.
The second kind of correspondence that can be called translation
is illustrated by English John, Spanish Juan, Russian Ivan, Italian
Giovanni, German Johannes, Scots Gaelic Ian, and so forth. The fact
that such names are parallel historical developments of a common
ON DEFINING THE PROPER NAME
prototype is not crucial, but the fact that they are used in parallel
ways and serve as equivalents, for example in biblical translation,
is decisive. Place names of the type are Vienna and Wien, Florence
and Firenze. Items that have no etymological connection but which
are functionally equivalent in the same way are Finland and
Suomi, Germany and Deutschland, Everest and Chomolungma. To
say that the first word in each set is the English version and not the
English translation of the others is to quibble over terminology (in
that we might as well say that dog is the English equivalent rather
than translation of Hund) or to argue in a circle (if we assume
names cannot be translated, then names that are apparently trans-
lational equivalents must be something else). Although equiva-
lences of this kind are different from calques, in which idiom forma-
tion is involved, both can reasonably be called translation.
The most common sort of equivalence, however, is one that has
seldom if ever been thought of as translation, but which has a good
claim to the title. When we use one language, such as English, and
wish to refer by name to a person, place, or other nameable thing
from a culture in which another language is used, and if there is
no readily available calque or equivalent of the kinds noted above,
the usual procedure is to find out the name in the native language
and then to render it in some approximate way in English. Such
rendering takes place in either of two forms. Suppose the foreign
name is Spanish Quixote [kix6te]. On the one hand, we may imitate
the foreign pronunciation in some reasonably close way, allowing
for differences between what sounds are found in each language
phonemically and phonetically and for phonotactic differences be-
tween the two languages. Thus the Spanish name may be rendered
into English [kiyh6wdi]. The first two vowels, simple in Spanish,
are diphthongized; the last vowel, being unstressed, has to be re-
placed by a different one in the English dialect considered here,
which permits final unstressed [i] but not [e]. The second con-
sonant, a velar fricative in Spanish, is replaced by its nearest English
equivalent, aspiration. The last consonant, a Spanish dental, is re-
placed by an English alveolar and further is voiced, since this
variety of English pronounces latter like ladder, atom like Adam,
and so in all similar environments. The process just described is
familiar to linguists under the term "sound substitution."
Alternatively, the spelling of the foreign name may be adapted
to English. In the case of Quixote, the Spanish spelling and its
SEMANTIC NAMES: DEGREE OF MEANING
English adaptation are identical; however, if the foreign language
uses a nonroman alphabet or a nonalphabetical writing system, such
as the Cyrillic alphabet or the kana syllabaries, the written form
will require change. In either case the spelling is then pronounced
in English according to the rules of English orthoepy. The name
Quixote so treated yields the pronunciation [kwfksat] or [kwfksbt].
This is the familiar process of giving a spelling pronunciation to
the transliteration of a foreign name.
Sound substitution is the phonological equivalent and translitera-
tion is the orthographic equivalent of translation. As translation is
the pairing of lexical and grammatical units in two languages, so
sound substitution is the pairing of phonological units and trans-
literation the pairing of orthographic units. In themselves they are
not translation but devices by which a language borrows words,
whether names or other sorts of words, from a foreign tongue. Once
such a word has been borrowed, it becomes a part of the borrow-
ing language in phonology, morphology, and syntax, although it
may for some time continue to be marked as a "loan word" or
"foreign vocabulary" in a variety of ways. As part of the borrowing
language, it is available to serve as a translation of the correspond-
ing word, its etymon, in the native language. We do not usually
say that English Quixote, however pronounced, is a translation of
Spanish Quixote, but there is good reason for doing just that. The
English name can be regarded as a translation of the Spanish be-
cause the two have parallel uses, for instance when the former serves
as a translational equivalent for the latter in reference to the fictional
character from La Mancha. (On the similar but more specific
question of the naturalization of immigrants' surnames, see Dykema
The process just described whereby a name is borrowed, and
thereafter is available to serve as a translation of its etymon, is
not different from the same process applied to common words. Thus
the American soldier stationed in Japan acquires the Japanese
word for 'good-bye' by sound substitution as [sayonadl], and the
student of Indian religion acquires a Sanskrit philosophical term as
dharma [darma] by transliteration and spelling pronunciation.
Thereafter either term is available as an English translation for the
foreign equivalents. The loan word dharma is favored as a transla-
tion of the Sanskrit technical term because possible calques, such
as law, fail to cover adequately the semantic area of the Sanskrit
ON DEFINING THE PROPER NAME
word. The popular loan sayonada is a close equivalent to the older
English good-bye except in associations and circumstances of use;
it, however, is also an English word because it is used in English
contexts and conforms to all the rules of English. As such, it is
available for use as a translation of the Japanese word transliterated
sayonara, although the availability of the better established and
nearly synonymous good-bye may preempt it.
Dharma and sayonada are both recent loans, and the latter may
well not survive the generation among whom it originated, yet their
process of borrowing is not significantly different from that which
resulted in street or wine or episcopal at an earlier time. Nor is the
process that has resulted in any of those words significantly differ-
ent from the one that introduces proper names like Quixote into
English. The subsequent use of all such loan words as equivalents
for their etyma must be called translation, unless an arbitrary and
a priori terminological distinction is to be made. Consequently the
proposition that names are not translated cannot be used in support
of their meaninglessness.
To regard proper names as words with no linguistic meaning but
only extralinguistic reference suggests that they are not really part
of language. There are those who have maintained that names are
indeed not a part of language (Harris 1751:346; Strawson 1950:186;
Vendler 1967:38; McCawley 1968:138-39), but the idea is counter-
intuititive and smacks of the old dodge whereby the only language
facts that are recognized are those that can be conveniently handled
by the available theory. Utley's (1963:165) view that "a grammar
which does not include proper names is no grammar at all" has
more to recommend it.
Sometimes instead of holding that names themselves are not
part of language (which seems transparently false), the position is
taken that knowledge of names is not part of knowledge of a
language. So put, there is a sense in which the view is correct. A
speaker of English may never have encountered the name Blavatsky
or, having encountered it, may not know that it is applied to such
and such a person. He would not thereby cease to be a speaker of
English, and in that way it can be said that knowledge of names
is not knowledge of the language. However, what has just been
said about names is also true of appellatives. A speaker of English
may never have encountered the word blastophore or, having en-
countered it, may not know it applies to such and such a thing. He
SEMANTIC NAMES: DEGREE OF MEANING
would not cease to be a speaker of English for that reason either.
Complete knowledge of the vocabulary available to English speakers
is not a prerequisite to being an English speaker. It would certainly
be unusual if a speaker did not know the words blame or blanket
or blast, but it would also be unusual if he did not know Boston
or Blondie or Black Forest. The lack of such knowledge would be
debilitating in the easy use of English but would not prevent it
altogether. Paraphrases of various sorts could replace each of the
individual words. On the other hand, if a person knew no common
nouns at all or were unable to use common nouns appropriately in
sentences, it would certainly be said that he was not really a speaker
of English; and substantially the same thing could be said about a
person that knew no proper names or was unable to use any mem-
bers of the class in appropriate ways. Ignorance of either class as a
whole would be radical ignorance of the language system.
To say that ignorance of names, unlike ignorance of appellatives,
is not ignorance of the language may imply something like the fol-
lowing: X, who is a speaker of English, encounters Y on the street
while both are watching a building under construction. They have a
short conversation about the construction, the weather, and team-
standings in the National League. Then they part without having in-
troduced themselves. X does not know Y's name. Yet X is not
ignorant of the language. But then neither is he ignorant of names;
rather he is ignorant about Y, a nonlinguistic matter. If this is what
is intended, the catchphrase has been wrongly put and should
rather be "Ignorance of a person is not ignorance of a language,"
which is true enough but is not a statement about names. In no
sense is it true that there is some special way in which names can
be considered not part of language.
The assumption that names are words with reference but no
meaning can lead to a number of corollaries: that they are like
variable indices or pronouns, that they are applied arbitrarily, that
they are untranslatable, and that they are not part of a language
at all. These corollaries have unacceptable consequences in ordinary
language, and so Mill's view must be rejected as inadequate for
In a position that seems to be directly opposite Mill's, Sweet
(1891:1.59), Breal (1897:183), and Jespersen (1922:438, 1924:65-
66, 71) have argued that since proper names are more specific in
reference than common nouns, they must connote more attributes
ON DEFINING THE PROPER NAME
and be therefore the most meaningful of all words. The fewer things
a word stands for, the more specialized it is and the fuller its def-
inition must be in order to circumscribe the few objects, or in the
case of proper names "of the good old kind" the single object, of
reference. The views of Mill and of Sweet and the others are, how-
ever, not so much opposites as different statements of the same
position that seem contrary because they use the same words, mean-
ing or connotation, in different ways. Thus, Mill and Jespersen
both talk about connotation but understand it differently. Mill in-
tended a minimal set of defining characteristics, whereas Jespersen
seems to have had in mind the totality of a word's associations. If
we take Frege's (1892:56) three-way distinction between referent
(the thing to which a sign refers), sense (the conditions under
which a sign may appropriately refer to a thing), and association
(ideas arising in a speaker's mind from memories of sense impres-
sions), for Mill connotation was 'sense' and for Jespersen 'associa-
tion,' a variation in use of the term that still exists. This is not to
suggest, however, that Jespersen and the others were unaware of the
distinction; Jespersen was familiar with it but denied that any sharp
line can be drawn between the essential characteristics (sense) of
a word and the accidental ones (association).
Because Jespersen did not distinguish between a word's basic
sense and its associations, the question of how it is that proper names
like Quisling, Silhouette, Moll, Macintosh, or China become com-
mon appellatives seemed to him to be a crucial one. He thought that
in making them common words, we focus on one part of their mean-
ing-which requires that they should have more meaning as proper
names than as appellatives. The question is certainly important, as
is also the opposite one of how common appellatives like charity,
red, smith, or long island come to be used as proper names, and how
they differ from the pseudo-proper use of words like Father,
Mother, and Teacher. But it is not clear that any of these questions
have the importance for the definition of name that Jespersen as-
sumed. Semantic change is not limited to narrowing and widen-
ing the scope of a word's meaning; buxom did not acquire the sense
'busty' nor tiger the sense 'aggressively virile male' through generali-
zation or specialization, but rather through a process of transfer
that involves the associations of the words. There is no problem
when a proper name like Martinet is taken to mean something
associated with one of the persons who has borne it and thereby
SEMANTIC NAMES: DEGREE OF MEANING
becomes an appellative, nor need such change of meaning imply
that the name Martinet is more meaningful than the appellative
One of the problems with the Sweet-Jespersen view is that it
makes proper names impossible to define in the way appellatives are,
that is, by specifying those characteristics of things by virtue of
which they can be referred to by the word. It would be necessary,
in effect, to give a complete description of the thing, a practical
impossibility with odd consequences (Searle 1958:169). Worse,
however, is the fact that this approach, like that of Mill, identifies
the meaning of a proper name with the thing or things it refers to.
Mill makes the identification openly by allowing proper names no
connotation but only denotation; Sweet and Jespersen make it
covertly by equating the meaning of a name like Plato with the
totality of characteristics by which the person Plato is distinguished
from all other persons. In both cases the meaning of a word has
been made coterminous with a particular thing, and thus, as
suggested in the last chapter, linguistic and nonlinguistic facts are
A third position, and a tertium quid between that of Mill and
that of Sweet and Jespersen, is the one proposed by Frege, in which
it is supposed, as in the others, that proper names refer to single
unique objects but in which it is held, with Sweet and against
Mill, that names have connotative meaning. The problem implicit in
Sweet's view, that the sense of a name threatens to become in-
finite by embracing every fact by which the person named is dif-
ferent from every other person, is avoided through the expedient
of supposing that the sense of the name is limited to some manage-
able set of facts that may differ for every user of the name:
In the case of an actual proper name such as 'Aristotle'
opinions as to the sense may differ. It might, for instance, be
taken to be the following: the pupil of Plato and teacher of
Alexander the Great. Anybody who does this will attach
another sense to the sentence 'Aristotle was born in Stagira'
than will a man who takes as the sense of the name: the
teacher of Alexander the Great who was born in Stagira.
So long as the reference remains the same, such variations of
sense may be tolerated, although they are to be avoided in
the theoretical structure of a demonstrative science and ought
not to occur in a perfect language. (Frege 1892:57-58)
ON DEFINING THE PROPER NAME
Frege's view is echoed in Russell's (1905, 1918) proposal to re-
place the proper names of ordinary language with descriptions:
"When we ask whether Homer existed, we are using the word
'Homer' as an abbreviated description: we may replace it by (say)
'the author of the Iliad and the Odyssey (Russell 1919:179). The
view has been argued also by Searle (1958, 1969) and by S0rensen
One of the problems with the intermediate view is that the mean-
ing of a name is made a wholly private matter. Each name can
and probably does have a different meaning for every person who
uses it; there is no public definition for any name. Moreover, there
is a different set of such unlimitedly numerous private meanings for
every person or thing to which a name is attached. Furthermore, for
any speaker who knows two names for the same individual, such as
Cicero and Tully or Stalin and Dzugashvili, those names would ap-
parently be synonyms, thus making statements of the form "Dzu-
gashvili was Stalin" analytically true for that speaker, a position
accepted by Searle (1958). Similarly, Frege was apparently willing
to accept as a consequence that the statement "Aristotle was born
in Stagira" would be analytically true for anyone who takes as the
definition of Aristotle 'teacher of Alexander the Great who was born
in Stagira' but not for one who takes as its definition 'pupil of
Plato and teacher of Alexander.' These are odd consequences.
John Searle (1958), in supporting Frege's view, has argued that
using a name to refer to an object presupposes the truth of a set of
uniquely referring descriptive statements that are regarded by
users of the name as essential and established facts about the
object. It is the nature of this set of statements, however, that it
should be open, since what is regarded as basic established fact
about an individual is subject to easy revision; it is at any given time
unspecified, and if we should be required to specify the set, the
decision as to what facts belong to it would have to be made arbi-
trarily. Searle seems prepared to accept these consequences, but
for the grammarian they vitiate any usefulness in the proposal. A
meaning that consists of an indefinite, unknown, and arbitrary set
of statements is no meaning worth bothering with, for it cannot be
defined and is beyond agreement or disagreement. Furthermore, if
the sense of a proper name were always an imprecise one (Searle
1969:140), then we should be in doubt about how to apply the
name. But that is not the case.
SEMANTIC NAMES: DEGREE OF MEANING
In addition to its own special difficulties, this view of the meaning
of names shares with those of Mill and Sweet a confusion of
langue and parole, of the system of language with its application.
Consider two putative definitions of the expression my son: (1) 'first-
degree male descendent of the speaker' and (2) 'twelve-year-old
(in 1972) residing with the author of this study.' The former is a
possible definition for linguistic study in that it is applicable to
any correct, literal use of the expression. The latter is not a definition
at all, but a characterization of one possible referent of the expres-
sion. The former, an intensional definition, is a statement about
the expression's meaning in the English language; the latter, an
extensional definition, is a description of one of the objects of which
the expression might be used. The Mill-Sweet-Frege approach to
defining proper names deals exclusively with extensional definitions.
That approach is doubtless useful for logic but has no application
whatever to the meaning of names in a linguistic sense, because
extensional definitions are about things rather than about language.
The problem of defining proper names is intransigent enough
to have led Jespersen (1924:70-71) to the conclusion "that no sharp
line can be drawn between proper and common names, the differ-
ence being one of degree rather than of kind," a view shared
by Breal (1897:182-83), Pulgram (1954), and Manczak (1968a:
218). But such a throwing up of hands is unsatisfactory, if for no
other reason, because it means that onomastics would be without a
clearly definable object of study-a consequence not dear to the
heart of onomatologists. But more important, the conclusion is un-
satisfactory because ordinary language implies, and twenty-four
hundred years of scholarship has assumed, that all languages have
names that are somehow clearly different from other sorts of words.
A theory of names that fails either to take account of those facts or
to define its object satisfactorily is inadequate.
Semantic Names: The Kind of Meaning
IF THE amount of meaning a name has, whether none
or infinite or some unspecified intermediate quantity,
is not an adequate characterization of the class, the kind of mean-
ing may be distinctive. It will be helpful, therefore, to look at some
of the ways a name can be used to see whether there is a clue in
them to a more adequate definition of the term. In each of the fol-
lowing sentences, Thomas is used in a different sense:
1. His name is Thomas.
2. Here are Catherine and Thomas.
3. She was a Thomas before she married.
4. I remember a younger Thomas.
5. He is trying to be another Thomas.
6. Looking in the mirror, he saw a Thomas looking back.
7. The dress she bought is a genuine Thomas.
In the first sentence, Thomas is really a citation form and might
more properly have been written in italics: "His name is Thomas."
The sentence in which it occurs is a metalinguistic one, being
language about language. Such uses must be set apart, because in
them the referent of the word is the linguistic form of which the
token is an exponent. The name in the other sentences might be
glossed as follows:
2. '(the) person (I intended and expect you to identify)
3. 'member of the Thomas family (i.e., family called
4. 'particular aspect of Thomas (i.e., person called Thomas)'
SEMANTIC NAMES: KIND OF MEANING
5. 'person with some characteristics of a particular Thomas
(i.e., person called Thomas)'
6. 'image of Thomas (i.e., person called Thomas)'
7. 'product of Thomas (i.e., person called Thomas)'
What all the glosses have in common is that they include either
the word Thomas as a citation form or a term that must itself be
so glossed. The examples therefore suggest that a proper name may
be defined as a word whose definition includes a citation of the word
itself. In the gloss for sentence (2), the citation form occurs as an
element in a differentiating clause that directly modifies the genus,
whereas in sentences (3) through (7) the citation form is an
element in a clause that glosses some modifier of the genus. Names
of the former kind can be called primary proper names, and
those of the latter kind secondary proper names. Between these two
sorts of names there are a number of differences in addition to the
form of their glosses.
The bestowal of a primary proper name on a referent is ad hoc,
in that there is no way to predict what primary name may be given
to any particular referent. Consequently there is no way to know the
correct use of primary names apart from observing instances of their
use. Secondary names are bestowed according to rule and con-
sequently their correct use does not depend on such observation. A
speaker can know that a person is named Catherine only if he has
heard or seen the name used in connection with her, and the initial
decision to call her Catherine was an act of name-giving. On the
other hand, if a speaker knows the name of the family to which
an individual belongs, he will know what last name the individual
bears without ever having heard or seen it used of the individual,
and the initial application of the family name to the individual was
not an act of name-giving at all because it follows automatically
from general rules in our society. A family name is shared by the
members of a family group and is regularly changed only by legal
procedure, such as adoption or the marriage of females, as a con-
comitant of change of family membership. A family name is there-
fore primarily the name of the family group and only secondarily
the name of an individual member.
Many secondary names have a tendency to pass easily into appel-
latives, so that whether a given use is proper or appellative may not
ON DEFINING THE PROPER NAME
be immediately clear. If we say, "Edinburgh was the Athens of
Scotland," Athens is used as a secondary proper name meaning
'place with certain characteristics of Athens'; it will generally be
understood as metaphorical, intended to call the original Athens to
mind. However, if we say, "He is attending an academy," a metaphor
is unlikely; the connection with the proper name Academy has
faded, leaving the word as a pure appellative. Words continually
move from the status of proper names, via secondary use, to the
status of appellatives (see Partridge 1950 for examples). There may
be doubt about the standing of a word that is in the process, for
example Casanova, Mata Hari, or Valentino, and it is quite possible
that for some speakers Bluebeard is a secondary proper name, de-
rived from the Perrault character, while for others it is a simple
appellative meaning 'uxoricide.'
The distinction between primary and secondary names is as
"delicate" (in Halliday's sense) as it seems advisable to make. It
would be easily possible to recognize more remote derivations. Thus
Israel 'country called Israel is a primary name; Israeli 'citizen of
Israel (i.e., country called Israel)' is a secondary name; and sabra
'native-born Israeli (i.e., citizen of Israel (i.e., country called
Israel))' would then be a tertiary name. Nothing prevents names
being quarternary, quinary, and so on; however, the distinctions
cease to have any practical significance. In common usage, primary
and usually also secondary names are felt to be proper; tertiary
(and higher degree derivations) are not. For practical purposes, a
proper name can be said to be a word like Thomas, whose mean-
ing is 'person called Thomas,' or like Chicagoan, whose meaning
is 'inhabitant of the place called Chicago.' More generally, a proper
name is primarily any word X whose meaning can be expressed as
'entity called X' and secondarily any whose immediate definientia
include a term with such a meaning.
It may be objected that this schema for defining names is circular
or involves infinite regression, on the grounds that the definition in-
cludes the term being defined (Br0ndal 1948:60; Searle 1969:139).
But the grounds are illusory and the objection invalid. The illusion
of circularity is created by the appearance of the citation form X-
whatever it may be, Thomas, for example-both as the definiendum
and as part of the definiens. An examination of what is meant by
"citation form" is therefore in order, followed by a more careful
restatement of the definition schema for names.
SEMANTIC NAMES: KIND OF MEANING
Citation forms are used in a variety of ways. For example, we
can say either "Cat is spelled with three letters" or "Cat has three
letters," using the two statements interchangeably. However con-
venient it is to vary one's expression in this way, the two statements
are not equivalent, strictly speaking, because the citation form cat
stands for different things in the two uses. The first statement might
be paraphrased as "J cat is spelled with three letters," that is, the
lexical formative or morpheme is represented by a three-letter
spelling; the second statement as " has three letters," that is,
the orthographic sequence is composed of three letters. The subjects
are not interchangeable: "Jcat has three letters" is nonsense be-
cause morphemes are not composed of letters. We can also say
things like "Cat is pronounced /ket/," "Cat is a noun," "Cat means
'domestic feline,'" or "Cat is the genus 'feline' delimited by the dif-
ferentia 'domestic,'" or any combination of such statements. By
making all these statements about the same citation form cat, we
seem to be saying that they are all true of the same entity. They are
not. The citation form cat is a covering label, a fudge form, that can
stand for any of several entities existing on different linguistic
strata. To speak more precisely, we would have to make statements
like the following:
, which is the spelling for the morpheme I cat and
represents the pronunciation /kaet/, has three letters.
/koet/, which is the pronunciation of the morpheme Jcat
and can be spelled , consists of three phonemes.
~cat) is a noun, is pronounced /koet/, is spelled ,
and means 'domestic feline.'
'cat' is a shorthand way of expressing the genus 'feline'
and the differentia 'domestic,' which is the meaning of
Normally, there is no harm in using the citation form cat as a
cover term for any of , /ket/, cat or 'cat.' It is convenient
typographically and lets us be looser with our wording than we
would have to be if we were expressing ideas with greater exactness.
If, however, we are misled by this loose practice to hypostatize a
single entity behind the cover term, mischief may result. One
possible bit of mischief is that the definition schema proposed above
ON DEFINING THE PROPER NAME
may be thought of as circular. If, however, we replace the citation
form in that schema with more precise expressions, we have the
The definition of any proper-name morpheme (sequence)
JX} is 'entity (with such other differentiae as are appropri-
ate) referred to in speech as /Y/ and in writing as .'
A specific example of the schema is the following:
JThomas means 'person referred to in speech as /tAmas/
and in writing as .'
When the ambiguity represented by the citation form is eliminated
from the definition, the vicious circularity disappears too.
It may also be objected that this approach to defining the proper
name does not distinguish it from a common appellative (Ullmann
1952; S0rensen 1963:23). That is, if Bill means 'entity called Bill,'
it is proposed that boy similarly means 'entity called boy.' But there
is an important difference between the two sorts of words, and
their meanings are not parallel. Suppose a strange animal wanders
into my garden from the wilderness and I call out to my wife in
the house, "There is a beast in our garden." Then she might ask,
'"Well, is it a cat, or an elephant, or a unicorn?" and I could
reply, "It's a cat" or "None of those-it's an alligator." Though I
had never seen the particular beast before, I would know whether
it is or is not a cat. However, if my wife should ask instead, "Well,
is it Macavity, or Pyewacket, or Caligula?" I would have to answer
something like, "How should I know-we haven't been introduced
yet." To know that a creature is appropriately referred to by the
word cat, it is not necessary to observe anyone calling it "cat."
But to know that the same creature is appropriately called Pye-
wacket, it is necessary to observe some instance of the use of that
name with reference to the creature.
It is necessary to distinguish between the bestowal of a name
and its subsequent use, a distinction that is not the same as that
between the invention of an appellative neologism and its sub-
sequent use. The invention (which can be taken here as including
borrowing) of a new common word like googol or sputnik is
parallel to the invention (or borrowing) of a new name like
Mauvenia or Rukmini. The use of the appellative sputnik to refer
SEMANTIC NAMES: KIND OF MEANING
appropriately to a particular object (an artificial satellite) is
parallel to the use of Rukmini to refer appropriately to a particular
person (a woman named Rukmini). But with names there is an
event other than invention or use, an event that is not paralleled
by anything connected with appellatives. Names are bestowed on
particular individuals, and for a name to be associated with an
individual, it is necessary that an act of bestowal should have
taken place (Hockett 1958:311). Appellatives are invented and
used; names are invented, bestowed, and used. The bestowal of a
name may be deliberate, though informal, as when parents con-
sider what name to give a new child; it may involve elaborate and
extensive study, as when a public relations company gives a new
name to a movie star; it may be off-hand, as when a child casually
names a pet turtle. But in every case there is a specific act by
which the name is first assigned to the bearer. There is no such
act for appellatives. Appellatives are used but not bestowed; use
of a name presupposes an act of bestowal. Thus the definition of a
proper name X as 'entity called X' specifies not only a sufficient, but
also a necessary condition for the correct application of the name.
If the creature in my garden is a domestic feline, I will know
that it is appropriately referred to by the word cat even though
no one has ever before so referred to it. I cannot under similar
circumstances know that it is to be called Pyewacket.
It is presumably this characteristic of the application of names
that led Mill to conclude they are connotatively meaningless. How-
ever, Mill's conclusion goes too far by unnecessarily restricting what
is to be accounted meaningful. What we are to understand by the
term meaning is a vexed question. Gilbert Ryle (1957:128) has
observed that "preoccupation with the theory of meaning could be
described as the occupational disease of twentieth-century Anglo-
Saxon and Austrian philosophy." It would be hubristic here to
suggest any cure for the disease. But if we say that a word's meaning
is a statement of the conditions necessary for its appropriate use, or
as Ryle (1957:143) puts it, "to know what an expression means . .
is to know the rules of the employment of that expression," then no
word can be meaningless. Just as part of one meaning of cat is
that any possible referent be a domestic feline, so part of the
meaning of Pyewacket is that any possible referent be in fact
called Pyewacket. To be a cat it is not required that a creature
ever in fact be so called; to be a Pyewacket, it is. Thus names and
ON DEFINING THE PROPER NAME
appellatives alike are meaningful words in the vocabulary, although
their kinds of meaning differ.
The significant contrast between names and appellatives is not to
be overlooked. In addition to the differences in the content of
their definitions, names, unlike appellatives, are freely and imagina-
tively created de novo, as the colorful examples of personal names
collected by Thomas Pyles (1947, 1959) attest. Moreover, the
bestowal of names, whether traditional or innovative, is a pre-
rogative of every speaker blessed with the possession of children,
pets, boats, or other such nameables. It is hard for anyone to in-
vent a new appellative and get his application of it accepted by
others; but when it comes to names, every citizen is the equal of
the primordial Name-giver of Plato's Cratylus. Also unlike the proc-
ess of associating an appellative with a thing, the giving of names
often involves a formal ceremony, a ritual or legal imposition, as in
christenings, registrations, and launching, which regularly include
a performative utterance by which the name is made official:
"Name this child. William Henry." Indeed appellatives are not
"given" at all; they belong to a thing automatically by virtue of
other characteristics it possesses. Names, however, are bestowed
by someone, and only after that initial use in the act of bestowal do
they become a characteristic of the subject to be observed by
others and thus imitated.
It is also possible that names are governed by laws, synchronic
and diachronic, that are different from those applying to appellatives
(Manczak 1968a, 1968b). Moreover, the name-stock of any language
includes potentially the names of all other languages. Whereas
appellatives are only exceptionally borrowed, the reverse is true
for names. Names translated either as calques like Red Square and
Arch of Triumph or as established native equivalents like Leo
for Lev (Nikolaevich Tolstoi) are comparative rarities, the normal
procedure being to accept the foreign name as a loan with whatever
phonological, orthographical, and grammatical adaptations the bor-
rowing language requires. So English has The Hague not The Port,
Vladivostok not Lordeast, and Juan Carlos not John Charles for the
Spanish royal person.
The grammar of names and appellatives is not altogether the
same. It was noted in chapter 3 that the morphology of names need
not follow that typical of nouns. In syntax, also, names have special
characteristics. Thus, they alone occur in various naming construc-
SEMANTIC NAMES: KIND OF MEANING
tions: a man named Ishmael, Ishmael by name, of the name Ishmael,
and so forth. The earlier history of such constructions in Indo-
European has been treated by Hahn (1969); something like them
may be presumed to occur in all languages. They were studied at
least as early as the fourteenth century, when Thomas of Erfurt
recognized a class of expressions with a verbum vocativum in his
modistic grammar (Bursill-Hall 1966:141). Some constructions of
the type seem to admit either proper names or appellatives: "They
called him Ishmael"; "They called him a rascal"; "They called him
a sailor." However, the superficially similar constructions have quite
different interpretations when a proper name and when an appel-
lative is used in them. Thus it is a fact that the referent is Ishmael
because he is so called; but it would be wrong to say (barring an
extreme nominalist stance) that he is a rascal or a sailor because
he is so called. In such constructions, names are used as citation
forms and might be italicized or put in quotation marks, but
appellatives do not normally have that interpretation; that is, in
"Call me Ishmael," the word Ishmael is "mentioned," but in "Call
me irresponsible," the word irresponsible is "used." Similarly, a
sentence like "This is Golda" is ambiguous between two meanings:
on the one hand, it may be an identification, with the sense 'This
is the person Golda (whom I presume you know about)'; on the
other, it may be an introduction, with the sense 'This person (whom
I do not presume you to have any prior knowledge of) is called
Golda.' With the first sense Golda is being used; with the second,
mentioned. Appellatives occur with "introduction" meanings only
in very restricted contexts, chiefly pedagogical, as when in a foreign
language class, an instructor says, "This is a kulero, and this is a
Such striking peculiarities mark names as a special kind of lexical
item, but hardly exclude them from the realm of linguistic fact. In-
deed, it may be said that there are three main kinds of vocables
in the lexicon: those words that form closed systems in the grammar,
such as the articles, pronouns, auxiliaries, and some prepositions and
conjunctions; second, appellative nouns and those verbs, adjectives,
and so forth that form open sets and constitute the bulk of words
listed in general dictionaries; and last, names, which are often ex-
cluded from dictionaries on either theoretical or practical grounds.
It is the contention of this study that there is no valid theoretical
basis for excluding names from a dictionary, although their vast
ON DEFINING THE PROPER NAME
numbers and the small amount of linguistic information appropriate
to each offer ample practical justification for leaving them out of
general dictionaries. When dictionaries do include names, they
tend to give encyclopedic rather than linguistic information about
them, that is, they tend to discuss prominent persons or places
named rather than the names themselves. There are exceptions, such
as the "Pronouncing Vocabulary of Common English Given Names"
appendix in the Merriam New Collegiate dictionaries. Fuller ex-
amples of appropriate lexicographical treatment of names are
specialized works like Withycombe (1950) and Smith (1956); an
early example is Walker (1818). On the other hand, a work like
Webster's Dictionary of Proper Names (Payton 1970) is actually
an encyclopedia of proper names, as are most works treating place
Names, whether considered lexically or referentially, are of
many kinds, of which names of persons are the most central. Alone
among nameable objects, persons are always assumed to have
names. Whereas places, astronomical bodies, and events may be
nameless, if we encounter an anonymous person, we ask about
his name and if it is inaccessible, like Robinson Crusoe in his
discovery of Friday, we supply the lack. Some entities, like animals
and supernatural beings, will be expected to have names just in
case they are treated as persons in other ways, such as requiring the
relative who rather than which. Personal names are central in
another way also. Although systems differ from culture to culture,
it is common for a person to bear both primary and secondary
names as part of a unified naming system. Other kinds of name-
able objects seldom have names of both kinds.
Personal names are of two main sorts: legal names and bynames.
Their kinds vary according to the naming system, some of whose
manifold variations are noted by Levi-Strauss (1966:172-216), but
include given names (Pollyanna, Timothy), which are primary
proper names, and secondary proper names such as patronymics
(Thorkelsdottir, Ivanovich) and family or clan names (Jones,
Macintosh). Some naming systems distinguish between an autonym,
which is most like a given personal name in the English system, a
teknonym, meaning 'father or mother of so-and-so,' and a necronym,
which defines the individual's kinship to some deceased relative, for
example, Elder-brother-dead; one person may have several such
names that he bears serially according to the changing events of
SEMANTIC NAMES: KIND OF MEANING
his life (Levi-Strauss 1966:191-96). Bynames include hypocorisms
(Will, Bill, Jeannie-although such apparent pet names can also
occur as legal names), epithets used in addition to another name
(Stonewall Jackson, Jack the Ripper, Ethelred the Unready) or in
place of it (Hotspur, Old Baldy), code names (M, 007), noms de
guerre (Leclerc-as for the general Vicomte de Hautecloque of
the North African campaign), pen names (George Sand), stage
names (Engelbert Humperdinck), aliases adopted for concealment
(John Smith), and other pseudonyms (Atticus-as for Addison).
Although they are grammatically common, descriptive phrases like
the Iron Duke, the Blessed Virgin, the Little Corporal, the White
Queen, and the Lone Ranger may be conventionalized in a way that
makes them epithetic names. Bynames seem always to be primary
proper names, even when they are based on secondary names.
Thus the use of a hypocoristic form of a family name, such as
Jonesy, is not predictable in the way the family name itself is and
accordingly must be defined as a primary name.
A term for members of a group, organization, or community is
a secondary proper name because the definition of the term will
include reference to the name of the group. Thus names of national-
ities like Spaniard can be defined as 'native or citizen of Spain
(i.e., the country called Spain),' Angle as 'one of the Germanic tribe
of Angles (i.e., the group called Angles),' DeMolay as 'member of
the Order of DeMolay for Boys (i.e., the organization called
Order of DeMolay),' and mafioso as 'member of the Mafia (i.e.,
the organization called Mafia).'
Titles, as well as some kinship and occupational terms, are
regularly used with personal names, but are not themselves names
since they can be defined without reference to their signaling value.
Even when used alone as grammatically proper nouns, as in "Dad
is in the closet where Momma hung him," "Doctor will see you
now," or "Mister, you're next," such words are appellatives, not
names. To know that a woman can be called Madam, it is not
necessary to hear her so called.
The nonce use of expressions like sleepyhead, big spender, or
smartaleck as proper nouns, for example in "See whether you can
get sleepyhead to bed," does not imply that they are names, since
an appellative definition will adequately account for the pseudo-
name use. Personifications, as in Bunyan's "Wherefore Mercy began
thus to reply to her Neighbor Timorous," are problematical. They
ON DEFINING THE PROPER NAME
function grammatically in every way like proper nouns, but they are
probably best accounted appellatives rather than names, since the
personification has the designation it does not merely because it is
so called but because it is characterized by the quality signified by
A noteworthy item is Joe Doakes, which has two uses (apart from
any occurrence as a genuine personal name, not under consideration
here). First, it is used in the sense 'average, common man,' as in
"Joe Doakes pays the taxes and fights the wars." Second, it is used
with the sense 'person' of one whose name is unknown, as in "The
next Joe Doakes who cuts in front is going to get a surprise." What
is noteworthy about Joe Doakes is that it has all the associated
characteristics of a name: it is capitalized, is syntactially a proper
noun, and has the appearance of a hypocoristic given name plus
a family name. However, in neither of the uses noted above is it
a name. In both it is simply an appellative, as the glosses show.
Indeed, Joe Doakes is the very antithesis of a name, for we can be
reasonably confident that anyone of whom the term is used is
not in fact called Joe Doakes. Moreover, users do not need to have
observed an instance of its application to a person before knowing
that it can be applied to him; rather Joe Doakes is used of a par-
ticular individual precisely when the speaker has not observed any
use of his name, or it is used of a type, for which a name, in the
strict sense, is not possible. John Doe, Richard Roe, John Stiles,
and Richard Miles similarly are antinames, though with the added
complication that they are used in the legal register in the general
sense 'unnamed party to legal proceedings' and are further differenti-
ated as the first, second, third, and fourth such parties. In effect,
these terms constitute a special third-person pronoun system for
the category "proximity." Like Joe Doakes, they are not names.
Names for nonpersons are less interesting onomastically because
there is little variety in the kind of name, however much variety
there may be in the things named. Names are given to places:
eschatological realms (Eden, Lotus Land), astronomical bodies
(Aldebaran, Big Dipper, Venus), topographical features of many
kinds (Fujiyama, Mississippi, Old Faithful, Sahara, Golden Horn,
Heartbreak Ridge, Garden of the Gods, Gulf Stream, Micronesia,
Okefenokee), and political divisions (Nepal, Danelaw, Ulster,
Mohenjodaro, Cherokee Strip, Covent Garden, Portobello Road,
SEMANTIC NAMES: KIND OF MEANING
Although the difference between proper names and appellatives
is a discrete one, it is sometimes hard to decide whether a particular
item is a name or not. Thus Polaris is an appellative if it is defined
as the 'star most nearly aligned with the earth's axis over the north
pole,' in which case, 12,000 years from now, as a result of the earth's
precession, the star called Vega will be Polaris. On the other
hand, Polaris is a name if, like Regulus or Vega, it is simply what a
point of light is called. In the latter case, the fact that it was the
polar star at the time it was named is a historical fact rather than
a rule of usage.
Names are also given to historical events, whether their duration
is brief or lasting: Hegira, Children's Crusade, Glorious Revolution,
Fifth Republic, Age of Reason, and Cenozoic Era, for example.
Place names are often used for events: Vatican II, before Munich,
since Kent State. Many names of events are specialized uses of
appellative phrases. Thus, civil war, as a common appellative,
applies to any military conflict between sections or parties of the
same nation; but the phrase can become conventionalized as the
name for one or more particular historical events, such as the
English Civil War of the 1640s or the American Civil War of
the 1860s. The descriptive accuracy of the phrase is then no
longer apposite. Indeed, in the case of the American conflict, the
question of whether the war was a civil one or one between
sovereign states was the immediate issue over which the war
was fought. Despite the existence of alternative names of various
descriptive implications, such as Horace Greeley's Great Rebellion,
the United Daughters of the Confederacy's War Between the States,
and a score of others recorded by McDavid and McDavid (1969),
the appellative meaning of such expressions is largely irrelevant
to their use as names. As witness to that irrelevance, there is a
reference of the New Orleans Times-Picayune to the "Spanish War
Between the States," cited by the McDavids as an amusing example
of linguistic hyperdelicacy.
Whether Monday and other similar terms for recurring periods
in the calendar are proper names has been the subject of much
debate. Gardiner (1954:53-54), in a spirit of compromise, called
them "common proper names," but the implications of such a
coincidentia oppositorum are not altogether clear. If a definition
like 'second day of the week' is appropriate for Monday, it is an
appellative, as are the names of holidays, months, and seasons.
ON DEFINING THE PROPER NAME
Many other things receive names: institutions and organizations
(Royal Academy, Republican Party, General Motors, NAACP,
Yale, Dodgers), beliefs and practices (Islam, McCarthyism), animals
(Checkers, Moby Dick), plants (Yggdrasil, Bo Tree, General
Sherman Tree), and gems (Koh-i-noor, Cullinan). Generic terms,
whether English or Latin, popular or learned, are appellatives
(bald cypress, Taxodiaceae). Artifacts are named: ships (Pinafore),
trains (Twentieth Century Limited), aircraft (Spirit of St. Louis),
spacecraft (Odyssey), weapons (Hrunting), buildings (Alhambra,
Madison Square Garden, Kremlin), and various objects (Holy
Grail, Big Ben, Liberty Bell).
Works of art customarily have titles as a special kind of name:
books (War and Peace), periodicals (Time), comic strips (Pea-
nuts), musical pieces (Gaudeamus Igitur), documents (Tanaka
Memorial), plays (The Tempest), poems (The Seasons), paintings
(American Gothic), prayers (Angelus), statuary (Wingless Vic-
tory), films (The Great Dictator), television programs (All in the
Family), speeches (Gettysburg Address), and-although they are
perhaps not works of art-laws, treaties, and the like (Mann Act,
Names that are in fact used of only one referent, for example
Popocatepetl, offer a problem in the form of their definition. As a
name, the definition of Popocatepetl must be of the form 'entity
(or perhaps more specifically, volcano) called Popocatepetl'; but
because the word has unique reference, there is a possibility of
defining it extensionally as 'volcano near Mexico City, 17,887 feet
in height' or, more precisely, to include within the definition a
location by the coordinates of longitude and latitude. There are
practical difficulties in doing so. As Landau (1967:13) has observed,
"no present grid system, or locating technique, can tell us precisely
enough where most places on earth really are." Although we might
conclude that language is imprecise enough to use imprecise co-
ordinates, there are also theoretical objections to a definition that
implies the volcano has the name Popocatepetl because of its
Similarly it might be supposed that the name of a particular
person could be defined as referring to the individual begotten of
such-and-such parents, at such-and-such a time, as Tristram Shandy
was of Mr. and Mrs. Shandy in the ill-starred night between the
first Sunday and the first Monday in the month of March, in the year
SEMANTIC NAMES: KIND OF MEANING
of our Lord 1718. But the supposition is wrong, not merely be-
cause the same name can be given to more than one individual or
because of practical difficulties in ascertaining the time and agents
of begetting, but because so to define a name would imply that an
individual has the name he does by virtue of the circumstances of
his begetting. On the contrary, as is well-known, Tristram was so
named because he was so called by the Reverend Mr. Yorick, in
the face of the best parental intentions.
Extensional definitions, it has already been argued, are categori-
cally wrong for linguistic purposes. They deal not with language,
but with things. They are appropriate not for a grammar or
dictionary, but for an encyclopedia. Language is used to talk about
particulars; but within a language system, all terms are general
An important, but still unanswered, question is what constitutes
nameability, what properties a subject must possess in order to be
eligible for a name. The requirement most often suggested is that
the thing must be one that holds a special interest for speakers
(Christophersen 1939:63; Russell 1948:89; Gardiner 1954:45; Pul-
gram 1954:32; S0rensen 1963:105; Waismann 1965:198; Zabeeh
1968:67). Chomsky (1971:14) suggests that nameability depends on
the "function of an object in a space of human action," which is
perhaps only another way of talking about "interest." It is clearly
right that to be named, a thing must be something in which
speakers might be interested, but that is a necessary rather than
a sufficient condition. There are many things in which speakers
have a reasonably strong degree of interest, such as a lawn that
needs weekly mowing, a beefsteak on one's plate, or an antique
firescreen. Yet outside fantasy, where all is possible, it would be
exceedingly odd to call a lawn Filbert, a steak Hank, or a firescreen
Zoe. The problem is not that the particular names are incompatible
with their referents; it is rather that such referents do not get
named, no matter how much interest we may have in them.
Another obviously necessary (but also insufficient) requirement
is that the thing to be named must be recognizable as a gestalt.
"Thing-ness" does not, however, require spatiotemporal continuity.
As Chomsky (1971:14) points out, certain works of art, such as
mobiles, are nameable although they lack spatial continuity; and the
science-fictional James Kirk, transported from his spaceship to the
surface of an alien planet, remains James Kirk although continuity
ON DEFINING THE PROPER NAME
is broken (the counterfactual nature of the example is irrelevant
since the fiction is put forth not as fantasy, but as a reasonable
possibility and is accepted as such by the audience).
A number of other criteria have been proposed: that names are
given to entities so similar to one another that the differences be-
tween them are difficult to recognize or describe; that names are
useful for entities that can serve as reference points for identifying
other entities; that names, being economical means of referring,
are given to objects for which such economy is desirable; and that
names are given to things not susceptible to systematic ordering or
for which standardized order seems inappropriate (Gardiner 1954:
45; Shwayder 1963:57). None of the foregoing are quite satisfactory
criteria; and, indeed, it would be possible to argue for the exact
opposites of the first and last.
There are also two sufficient, although not necessary, conditions
for nameability: personhood and prominence. All persons are ex-
pected to have names; indeed an individual's name is bound up
with his personhood, a nameless creature being thereby deperson-
alized. Nonhumans that are personalized are also usually named,
for example, pets and supernatural powers. God's name may be
ineffable, but he has one; the absolute which is unnamed is also im-
Things that are prominent of their kind by virtue of intensity,
size, duration, complexity, or whatnot are likewise nameable. Thus
rivers have names, though the small creek flowing by my door has
none. The clock at Westminster is Big Ben, but that in the local
courthouse is anonymous. The Star of India is named, but not the
stone in my wife's ring. Hurricanes are Agnes, Beulah, and the
like, whereas tornados and summer storms are nameless. This
criterion does not fully explain why some things are named and
others not. Thus the event of 1666 when half the capital of
England burned down is called the Great Fire of London. The
event of 1945 when much of the city of Hiroshima was destroyed
has no distinctive English name, although it can be described by
various appellative phrases. How prominent a thing must be before
it is named is subjective and partly a matter of chance.
Although the decision of whether to name or not to name is
sometimes arbitrary or conventional, any subject that is either
a person or exceptionally prominent of its kind is regularly given
a name. And all named things may be presumed to be the objects of
SEMANTIC NAMES: KIND OF MEANING
human interest and to have a gestalt-like cohesiveness. Beyond
these criteria the requirements for nameability are not clear,
although they are matters of some interest and may reasonably be
supposed to be universals of human behavior.
The kind of definition that has been proposed here for proper
names is not wholly new. It is implicit in Mill's (1843:1.2.5) obser-
vation that "when we predicate of anything its proper name, when
we say, pointing to a man, this is Brown or Smith, or pointing to a
city, that is York, we do not, merely by so doing, convey to the
hearer any information about them, except that those are their
names." Although Mill did not draw it, the obvious conclusion would
seem to be that the information conveyed by a word is its meaning
and therefore the meaning of a name is that the referent is so
called. Marty (1908:438, 509) and Funke (1925:79) also noted this
implication of the use of names, but it was Gardiner (1954) who
recognized and most insistently called attention to the central role
played by the "distinctive sound" of a name in its use. Gardiner's
Theory of Proper Names is distorted by his almost exclusive
attention to "embodied" proper names, that is, names with a definite
referent in an act of parole-an ironic distortion in view of the
fact that Gardiner was one of the chief proponents of Saussure's
langue/parole distinction in England in the face of resistance to
such dichotomizing by his contemporary J. R. Firth, who thought of
it as "a quite unnecessary nuisance" (1957:144, 179, 227; Langen-
doen 1968:47). If we apply Gardiner's general theory to a con-
sideration of "disembodied" names as units in the system of
langue, a definition of the kind presented here follows naturally.
With respect to ordinary language, Gardiner's work on names is the
most important theoretical statement in recent times.
Another implicit recognition of the kind of meaning proposed
here for names can be seen in Strawson (1950:188): "What is ...
implied ... by my now referring to someone by name is simply the
existence of someone, now being referred to, who is conventionally
referred to by that name" (Strawson's italics). Although it is not
true that the use of a name implies the existence of the referent
(in the ordinary sense of existence), since fictional Mrs. Miniver
is as good a name as historical Mrs. Bracegirdle, Strawson is
certainly right that the referent may be assumed to be conven-
tionally referred to by the name; it is this fact that the definition
schema formalizes. Although Church (1956:5) did not himself
ON DEFINING THE PROPER NAME
limit the term so, he observed that "proper name is often restricted
to names . which have as part of their meaning that the denota-
tion is so called or is or was entitled to be so called." And according
to Ryle (1931:27), when we understand a proper name, "we
know that someone in particular is called by that name by
certain persons." Zink (1963:491) similarly combines a so-und-so-
genanntsein view of the meaning of names with an interpretation
requiring them to have unique reference, when he proposes that
"the meaning of a particular proper name 'P.N.' would be the
meaning of the expression 'the person truly named "P.N." who is
or was at the time T at place P.' Although he puts most emphasis
on the means by which a name is bestowed (as by christening or
the inheritance of a patronym), Shwayder (1964:453) argues the
use of a name "always implies that the speaker believes that the
indicated denotatum has been given that name."
Finally, Bach (1968:92) wants "to postulate that all nouns (at
least common nouns) are derived in one way, namely from
structures of roughly the form
Det + one + S
where S is further developed into a sentence of the form
Det + one + Aux + be + Predicate nominal"
and subsequently he notes that "within the framework suggested
here, names could be treated in two ways: They could be allowed
to occur as alternatives to variables within sentences, or they could
be derived from embedded sentences involving the predicate 'is
called or the like. If the latter course is followed, one
could deal with the semantic content of names in a way parallel to
that of other 'nouns' (121). The first alternative follows the tradi-
tion of assuming that names have unique referents, which was
argued against in chapter 4. The second, however, is the treat-
ment that is advanced in this chapter as the most accurate state-
ment of a name's meaning, in the sense of a rule that governs its
use. It is also in keeping with the logical tradition of Russell
(1905) and Quine (1950:220) that proposes to treat all singular
terms as descriptions and to reanalyze descriptions as predications;
in this tradition, too, all terms are general terms. Such a con-
SEMANTIC NAMES: KIND OF MEANING
vergence of disciplines in their view of language is a welcome
sign of the general usefulness of the analysis. Ordinary language and
the predicate calculus may have a resemblance on the semantic
The approach to defining proper names in this chapter has several
advantages that have not always been recognized:
1. It maintains that in respect to having meaning, proper names
are normal words in the lexicon rather than peculiar "meaningless"
words or extralinguistic phenomena. It normalizes proper names in
this way because it regards their referents as belonging to in-
tensional classes, like those of common nouns. In effect it says that
Thomases are individuals who can be recognized by the fact that
people apply the name Thomas to them, thus treating an individual's
name like any other attribute by which he can be recognized and
2. The form of the definition allows other attributes to be in-
corporated into it. If "A Boy Named Sue" is incongruous, the in-
congruity can be explained by a definition of the name Sue as a
'female person named Sue.' To the extent that names have connota-
tions (in Mill's sense) those connotations can be included in their
definitions, thereby also providing a partial explanation for the
belief that proper and common names differ by degree. The
necessary and sufficient condition for a proper name is that its
definition be of the form 'entity called X'; if other information is
included, for instance 'human' or 'female' or 'canine,' the name,
while remaining categorically proper, becomes more common-like in
3. The distinction between primary and secondary proper names
also formalizes the belief that proper and common names differ
by degree rather than categorically. The formal difference between
the meanings of primary names, like Delilah 'person called Delilah'
or Chicago 'place called Chicago,' and those of secondary names,
like a Delilah 'person with characteristics of a particular Delilah' or
Chicagoan 'inhabitant of Chicago,' explains why there has been
disagreement about whether items of the second kind are proper
names, appellatives, or some sort of semiproper name.
4. On the other hand, the apparent circularity of the definition
explains why Mill and others have thought names to be meaning-
less. If we are told we are to see a whale, we can confidently infer
some things about the creature-that it is aquatic, warm-blooded,
ON DEFINING THE PROPER NAME
and probably big. But if we are told we are to see Shome, all we
can be sure of is that the entity can be called Shome. That is the
meaning of the name, but it does not add much to our knowl-
edge of the name-bearer. And consequently it is easy to con-
clude that names have no meaning.
5. It has often been observed that users of a name must be
supposed to know some facts about the named object that could be
taken as the meaning of the name. But, as argued in chapter 4,
if we take the sort of facts that are usually suggested as illustra-
tions-that Aristotle was the student of Plato, or was the teacher
of Alexander, or was born at Stagira-we have to conclude either
that the meaning of a name is infinite, if we take the sum of those
facts, or that it is indeterminate, if we take any part of them.
There is, however, one fact about any referent that every namer of
the referent must surely be assumed to know: that the referent
is called by the name used in making reference; and it is this
generally known fact that the definition specifies. Thus the position
taken by those, like Frege, who believe a name to have some
sense is also justified, but the definition makes that sense definite
and describable, whereas otherwise it is not.
6. The definition treats names as an autonomous and purely
intralanguage semantic category. On the one hand, it escapes the
problems that come from trying to base a definition on extra-
linguistic reference; and on the other hand, it does not assume that
the semantic class of names must be isomorphic with any gram-
matical, phonological, or orthographic classes. English has capital-
ized versus lowercase words in its orthography, proper versus
common nouns in its grammar, and names versus appellatives in
its semantics, but the various classes, though partly correlate, are
independent of one another. In general, names are proper nouns
and are capitalized, but exceptions are easy to find. In respect to
naming, language would seem to be a multileveled and polysystemic
7. Finally, the definition has, it would seem, a fair chance of
being a universal of language. It is highly probable that all lan-
guages have words for which a necessary defining characteristic is
that such a word can be applied to an entity just in case the
entity is actually called by the word. Efforts to define a name
orthographically, phonologically, or grammatically will end in
language-specific phenomena that are of limited interest to the
SEMANTIC NAMES: KIND OF MEANING 87
general theory of onomastics; the name as a universal is a semantic
A name is therefore a word people use to call someone or some-
thing by. So we may conclude that Hermogenes was right after all,
and Cratylus wrong. If people called him Hermogenes, that was his
name, because what a man is called is, after all, what a name is.
Ashley, Leonard R. N. 1971. "Changing Times and Changing Names: Reasons,
Regulations, and Rights." Names 19:167-87.
Austin, J. L. 1961. "The Meaning of a Word." In Philosophy and Ordinary
Language, ed. Charles E. Caton, pp. 1-21. Urbana: University of Illinois
Ayer, A. J. 1963. The Concept of a Person, and Other Essays. New York: St.
Bach, Emmon. 1968. "Nouns and Noun Phrases." In Universals in Linguistic
Theory, eds. Emmon Bach and Robert T. Harms, pp. 90-122. New York:
Berle, Alf K., and de Camp, L. Sprague. 1959. Inventions, Patents, and Their
Management. Princeton, N.J.: Van Nostrand.
Bloomfield, Leonard. 1933. Language. New York: Holt.
Bolinger, Dwight. 1971. The Phrasal Verb in English. Cambridge: Harvard
1972. "Accent is Predictable (If You're a Mind-Reader)." Language
Bowman, William Dodgson. 1931. The Story of Surnames. London: Routledge.
Breal, Michel Jules A. 1897. Essai de semantique-science des significations.
Br0ndal, Viggo. 1948. Les Parties du discours: dtudes sur les categories linguis-
tiques. Traduction francaise par Pierre Naert. Copenhagen: Einar Munks-
Brown, Goold. 1851. The Grammar of English Grammars. Reprint. New York:
William Wood, 1873.
Bursill-Hall, Geoffrey. 1966. "Aspects of Modistic Grammar." In Report of the
Seventeenth Annual Round Table Meeting of Linguistics and Language
Studies, ed. Francis P. Dinneen, pp. 133-48. Monograph Series on Lan-
guages and Linguistics, no. 19. Washington: Georgetown University Press.
Carroll, Lewis. 1871. Through the Looking-Glass, and What Alice Found There.
In The Annotated Alice, ed. Martin Gardner, pp. 166-345. New York:
Clarkson N. Potter, 1960.
Chafe, Wallace L. 1970. Meaning and the Structure of Language. Chicago:
University of Chicago Press.
Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge: MIT
1971. Problems of Knowledge and Freedom. New York: Random
Christophersen, Paul. 1939. The Articles: A Study of Their Use in English.
Copenhagen: Einar Munksgaard.
ON DEFINING THE PROPER NAME
Church, Alonzo. 1956. Introduction to Mathematical Logic. Princeton: Prince-
ton University Press.
Collinson, William Edward. 1937. Indication: A Study of Demonstratives,
Articles, and other 'Indicaters.' Language Monographs, no. 17. Baltimore:
Coseriu, Eugenio. 1967. "El plural en los nombres proprios." In Teoria del
lenguaje y linguistica general: cinco studios, pp. 261-81. 2d ed. Madrid:
Curme, George 0. 1935. Parts of Speech and Accidence. Boston: Heath.
Dallas, James. 1913. "The Honorific 'The.'" Scottish Historical Review 10:39-
46. Comment by Herbert Maxwell, ibid., pp. 230-31.
Davidson, Thomas. 1874. "The Grammar of Dionysios Thrax." Journal of
Speculative Philosophy 8:326-39.
DeCamp, David. 1967. "African Day-Names in Jamaica." Language 43:139-47.
Der-Houssikian, Haig. 1970. "Definiteness and Indefiniteness." Manuscript.
Dillard, J. L. 1968. "On the Grammar of Afro-American Naming Practices."
Donald, Henderson H. 1952. The Negro Freedman. New York: Schuman.
Dykema, Karl W. 1972. "Anglicized Forms of Proper Names: A Report on a
Questionnaire Filled Out by Students at Youngstown State University." In
Studies in Linguistics in Honor of Raven I. McDavid, Jr., ed. Lawrence M.
Davis, pp. 359-65. University, Ala.: University of Alabama Press.
Feldman, Harold. 1959. "The Problem of Personal Names as a Universal Ele-
ment in Culture." American Imago 16:237-50.
Firth, J. R. 1957. Papers in Linguistics, 1934-1951. London: Oxford Univer-
Frege, Gottlob. 1892. "On Sense and Reference." In Translations from the
Philosophical Writings of Gottlob Frege, eds. Peter Geach and Max Black,
pp. 56-78. Oxford: Blackwell.
Funke, 0. 1925. "Zum Definition des Begriffes 'Eigenname. In Probleme der
englischen Sprache und Kultur, Festschrift fir Johannes Hoops. Germanische
Bibliothek, Abt. 2, Bd. 20.
Gardiner, Alan. 1951. The Theory of Speech and Language. 2d ed. Oxford:
-- 1954. The Theory of Proper Names. 2d ed. London: Oxford Univer-
sity Press. First printed in 1940.
Gleason, H. A., Jr. 1965. Linguistics and English Grammar. New York: Holt.
Gove, Philip B. 1965. "The Nonlexical and the Encyclopedic." Names 13:103-
Green, William. 1972. "Humorous Characters and Attributive Names in Shake-
speare's Plays." Names 20:157-65.
Hacking, Ian. 1968. "A Language without Particulars." Mind 77:168-85.
Hahn, E. Adelaide. 1969. Naming-Constructions in Some Indo-European Lan-
guages. Cleveland: Case Western Reserve University Press.
Halliday, M. A. K. 1961. "Categories of the Theory of Grammar." Word
1967-68. "Notes on Transitivity and Theme in English." Journal of
Linguistics 3:37-81, 199-244; 4:179-215.
Hamp, Eric P. 1956. Review of Theory of Names, by Ernst Pulgram. Romance
Harms, Robert T. 1968. Introduction to Phonological Theory. Englewood Cliffs,
Harris, James. 1751. Hermes; or, A Philosophical Inquiry Concerning Lan-
guage and Universal Grammar. Reprint. Menston, Yorkshire: Scolar Press,
Hill, Archibald A. 1958. Introduction to Linguistic Structures. New York: Har-
-- 1966. "A Re-Examination of the English Articles." In Report of the
Seventeenth Annual Round Table Meeting on Linguistics and Language
Studies, ed. Francis P. Dinneen, pp. 217-31. Monograph Series on Lan-
guages and Linguistics, no. 19. Washington: Georgetown University Press.
Hockett, Charles F. 1958. A Course in Modem Linguistics. New York: Mac-
Householder, Fred W. 1971. Linguistic Speculations. Cambridge: At the Uni-
Jacobs, Roderick A., and Rosenbaum, Peter S. 1968. English Transformational
Grammar. Waltham, Mass.: Blaisdell.
Jakobson, Roman. 1972. "Verbal Communication." Scientific American 227
Jespersen, Otto. 1909-49. A Moder English Grammar on Historical Principles.
7 vols. Copenhagen: Einar Munksgaard.
- 1922. Language: Its Nature, Development, and Origin. London: Allen
-- 1924. The Philosophy of Grammar. Reprint. New York: Norton, 1965.
Johnson, W. E. 1921. Logic. Reprint. New York: Dover, 1964.
Jowett, B. 1873. The Dialogues of Plato. 4 vols. New York: Scribner.
Keil, Heinrich. 1864. Grammatici Latini. 8 vols. Reprint. Hildesheim: Olm,
Krapp, George Philip. 1925. The English Language in America. New York:
Kruisinga, E. 1932. A Handbook of Present-Day English. 5th ed. 4 vols.
Lakoff, George. 1971. Foreword to Where the Rules Fail, by Ann Borkin et al.
Ann Arbor: Department of Linguistics, University of Michigan. Mimeo-
Lamb, Sydney M. 1966. Outline of Stratificational Grammar. Washington:
Georgetown University Press.
[Lancelot, Claude, and Arnauld, Antoine.] 1660. Grammaire general et
raisonnee. Reprint. Menston, Yorkshire: Scolar Press, 1968.
Landau, Robert M. (357-03-6623). 1967. "Name or Number-Which Shall
It Be?" Names 15:12-20.
Langendoen, D. Terence. 1968. The London School of Linguistics: A Study
of the Linguistic Theories of B. Malinowski and J. R. Firth. Research Mono-
graph, no. 46. Cambridge: MIT Press.
1970. Essentials of English Grammar. New York: Holt.
Leech, Geoffrey N. 1969. Towards a Semantic Description of English. London:
Levi-Strauss, Claude. 1966. The Savage Mind. Chicago: University of Chicago
Linsky, Leonard. 1963. "Reference and Referents." In Philosophy and Ordinary
Language, ed. Charles E. Caton, pp. 74-89. Urbana: University of Illinois
Long, Ralph B. 1961. The Sentence and its Parts. Chicago: University of
ON DEFINING THE PROPER NAME
-- 1969. "The Grammar of English Proper Names." Names 17:107-26.
Lyons, John. 1966. "Towards a 'Notional' Theory of the 'Parts of Speech.'"
Journal of Linguistics 2:209-36.
Manual of Style. 1969. 12th ed. Chicago: University of Chicago Press.
Manczak, Witold. 1968a. "Le Nom propre et le nom commun." Revue In-
ternationale d'Onomastique 20:205-18.
1968b. "Onomastik und Strukturalismus." Beitrige zur Namenfor-
Marty, Anton. 1908. Untersuchungen zur Grundlegung der allgemeinen Gram-
matik und Sprachphilosophie. Vol. 1. Halle: Max Niemeyer.
Matthews, C. M. 1966. English Surnames. London: Weidenfeld and Nicolson.
McCawley, James D. 1968. "The Role of Semantics in a Grammar." In Uni-
versals in Linguistic Theory, eds. Emmon Bach and Robert T. Harms, pp.
124-69. New York: Holt.
McDavid, Raven I., and McDavid, Virginia G. 1969. "The Late Unpleasant-
ness: Folk Names for the Civil War." Southern Speech Journal 34:194-204.
McMillan, James B. 1949. "Observations on American Place-Name Grammar."
American Speech 24:241-48.
Merton, Robert K. 1965. On the Shoulders of Giants: A Shandean Postscript.
New York: Harcourt.
Michael, Ian. 1970. English Grammatical Categories and the Tradition to 1800.
Cambridge: At the University Press.
Mill, John Stuart. 1843. A System of Logic. Reprint. New York: Harper, 1846.
Morris, Charles. 1946. Signs, Language, and Behavior. Reprint. New York:
George Braziller, 1955.
Mullally, Joseph P. 1945. The Summulae Logicales of Peter of Spain. Publica-
tions in Mediaeval Studies, no. 8. Notre Dame, Ind.: University of Notre
Miiller, F. Max. 1891. The Science of Language. Vol. 1. New York: Scribner.
Palmer, Harold E., and Blandford, F. G. 1969. A Grammar of Spoken English.
3d ed., rev. Roger Kingdon. Cambridge, England: Heffer.
Partridge, Eric. 1950. Name into Word: Proper Names that Have Become
Common Property. New York: Macmillan.
Payton, Geoffrey. 1970. Webster's Dictionary of Proper Names. Springfield,
Postal, Paul. 1966. "On So-Called 'Pronouns' in English." In Modern Studies
in English, eds. David A. Reibel and Sanford A. Schane. Englewood Cliffs,
N.J.: Prentice-Hall, 1969.
Poutsma, H. 1914. A Grammar of Late Modern English. 4 vols. Groningen:
Prenner, Manuel. 1942. "Ora Jones Married Ora Jones." American Speech
Pulgram, Ernst. 1954. Theory of Names. Berkeley: American Name Society.
Reprint from Beitrdge zur Namenforschung, vol. 5, no. 2.
Pyles, Thomas. 1947. "Onomastic Individualism in Oklahoma." American
1959. "Bible Belt Onomastics; or, Some Curiosities of Anti-Pedobaptist
Nomenclature." Names 7:84-100.
Quine, Willard Van Orman. 1950. Methods of Logic. New York: Holt.
1960. Word and Object. Cambridge: MIT Press.
1963. From a Logical Point of View. New York: Harper.
Quirk, Randolph; Greenbaum, Sidney; Leech, Geoffrey; and Svartvik, Jan.
1972. A Grammar of Contemporary English. London: Longmans.
Rennick, Robert M. 1970. "The Nazi Name Decrees of the Nineteen Thirties."
Roberts, Paul. 1967. Modern Grammar. New York: Harcourt.
Robins, R. H. 1966. "The Development of the Word Class System of the
European Grammatical Tradition." Foundations of Language 2:3-19.
1967. A Short History of Linguistics. Bloomington: Indiana University
Robinson, Fred C. 1972. "Appropriate Naming in English Literature." Names
Russell, Bertrand. 1905. "On Denoting." Mind 14:479-93.
1918. "The Philosophy of Logical Atomism." The Monist 28:509-27.
-- 1919. Introduction to Mathematical Philosophy. London: Allen and
1940. An Inquiry into Meaning and Truth. New York: Norton.
1948. Human Knowledge: Its Scope and Limits. New York: Simon
Ryle, Gilbert. 1931. "Systematically Misleading Expressions." In Logic and
Language, ed. Antony Flew, pp. 13-40. Garden City, N.Y.: Doubleday,
--- 1953. "Ordinary Language." In Philosophy and Ordinary Language,
ed. Charles E. Caton, pp. 108-27. Urbana: University of Illinois Press, 1963.
- 1957. "The Theory of Meaning." In Philosophy and Ordinary Lan-
guage, ed. Charles E. Caton, pp. 128-53. Urbana: University of Illinois
Sampson, Geoffrey. 1970. "Towards a Linguistic Theory of Reference." Type-
Searle, John R. 1958. "Proper Names." Mind 67:166-73.
1969. "The Problem of Proper Names." In Semantics-An Interdis-
ciplinary Reader in Philosophy, Linguistics, and Psychology, eds. Danny D.
Steinberg and Leon A. Jakobovits, pp. 134-41. Cambridge: At the Univer-
sity Press, 1971.
Shwayder, D. S. 1963. Modes of Referring and the Problem of Universals:
An Essay in Metaphysics. University of California Publications in Philosophy,
no. 35. Berkeley: University of California Press.
1964. Review of The Meaning of Proper Names, by H. S. S0rensen.
Journal of Philosophy 61:450-57.
Sledd, James. 1959. A Short Introduction to English Grammar. Chicago: Scott,
Sloat, Clarence. 1969. "Proper Nouns in English." Language 45:26-30.
Smith, Carlota S. 1964. "Determiners and Relative Clauses in a Generative
Grammar of English." Language 40:37-52.
Smith, Elsdon C. 1956. Dictionary of American Family Names. New York:
-- 1966. "The Significance of Name Study." In Proceedings of the Eighth
International Congress of Onomastic Sciences, ed. D. P. Blok. Janua Lin-
guarum, Series Major 17. The Hague: Mouton.
1969. "Influences in Change of Name." Onoma 14:158-64.
S0rensen, Holger Steen. 1958. Word-Classes in Modern English with Special
Reference to Proper Names. Copenhagen: Gad.
1963. The Meaning of Proper Names. Copenhagen: Gad.
Stageberg, Norman C. 1971. Introductory English Grammar. 2d ed. New York:
Stockwell, Robert P.; Schachter, Paul; and Partee, Barbara Hall. 1968. Inte-
ON DEFINING THE PROPER NAME
gration of Transformational Theories on English Syntax. Electronic Systems
Division USAF TR 68-419. Los Angeles: University of California.
Strang, Barbara M. H. 1968. Modem English Structure. 2d ed. New York:
St. Martin's. First published in 1962.
Strawson, P. F. 1950. "On Referring." In Philosophy and Ordinary Language,
ed. Charles E. Caton, pp. 162-93. Urbana: University of Illinois Press,
1959. Individuals: An Essay in Descriptive Metaphysics. London:
Sweet, Henry. 1891. A New English Grammar. 2 vols. Reprint. Oxford: Claren-
Togeby, Knud. 1951. Structure immanente de la langue frangaise. Travaux du
Cercle Linguistique de Copenhague, no. 6. Reprint. Paris: Larousse, 1965.
Ullmann, S. 1952. Review of Theory of Proper Names, by Alan Gardiner.
Archivum Linguisticum 4:66-67.
Utley, Francis Lee. 1963. "The Linguistic Component of Onomastics." Names
Varro, Marcus Terentius. 1964. De Lingua Latina, eds. G. Goetz and F.
Schoell. Amsterdam: Hakkert.
Vendler, Zeno. 1967. Linguistics in Philosophy. Ithaca: Cornell University Press.
Waismann, F. 1965. The Principles of Linguistic Philosophy, ed. R. Harre.
New York: St. Martin's.
Walker, John. 1818. A Key to the Classical Pronunciation of Greek, Latin, and
Scriptural Proper Names. New York: Collins and Hannay.
Wells, Rulon. 1954. "Meaning and Use." In Theory of Meaning, eds. Adrienne
and Keith Lehrer, pp. 113-35. Englewood Cliffs, N.J.: Prentice-Hall, 1970.
Withycombe, E. G. 1950. The Oxford Dictionary of English Christian Names.
Witkowski, Teodolius. 1964. Grundbegriffe der Namenkunde. Deutsche Aka-
demie der Wissenschaften zu Berlin Vortrige und Schriften, no. 91. Berlin:
Wolf-Rottkay, W. H. 1971. "Some Onomastic and Toponomastic Aspects of
Icelandic Traditionalism." Names 19:229-39.
Zabeeh, Farhang. 1968. What Is in a Name? An Inquiry into the Semantics
and Pragmatics of Proper Names. The Hague: Nijhoff.
Zandvoort, R. W. 1969. A Handbook of English Grammar. 5th ed. London:
Zink, Sidney. 1963. "The Meaning of Proper Names." Mind 72:481-99.
Zinkin, Vivian. 1969. "The Syntax of Place-Names." Names 17:181-98.