Skip to content
BY 4.0 license Open Access Published by De Gruyter Mouton March 5, 2021

Names of places

Katarzyna Kijania-Placek ORCID logo
From the journal Semiotica


The thesis advanced in this paper is that the proper names of cities or countries inherit the linguistic types of the nouns which denote the basic category of the objects the names refer to. As a result, in the case of the proper names of cities or countries, a reference by those names may select particular aspects of those objects, in the same way that book or newspaper selects the physical or informational aspects of objects in the extension of the nouns. This view is based on Asher’s and Pustejovsky’s conception of dot type semantics.

1 Introduction

The literature on proper names, including the recent debate between referentialism and predicativism, is largely informed by examples of the proper names of people, such as Aristotle or Alexander the Great. In this paper, I choose instead to concentrate on the names of places, particularly on the proper names of cities and countries. Some of them are syntactically distinguished by requiring the definite article, such as the Alps or the Netherlands, but this feature is idiosyncratic for English and thus not a feature of names as such.[1] The proper names of places share many features with personal names. Apart from individual uses exemplified by (1) and (2):

Alfred studies in Princeton. (Burge 1973: 429)
Britain lies under one metre of snow. (Cruse 2000: 117)

like personal proper names, proper names of cities and countries have systematic predicative uses:[2]

I’ve never met an Ophelia before. (Payne and Huddleston 2002: 521)
There are more than a dozen Salems in the United States. (Bach 2015: 772)

deferred uses (where the semantic value is an object other than the default referent of the term, see below for more constraints):

Obama left the Halloween party early.[3]
My parents protested during Vietnam.[4]

and descriptive uses (where the semantic value is a distributive property corresponding to the default referent of the term):

She’s no Florence Nightingale. (Payne and Huddleston 2002: 521)
Trieste is no Vienna.[5]

An interesting feature of the names of cities and countries, however, is that in contrast to the names of people, whose default referents are their contextually distinguished human bearers, the names of cities and countries systematically refer to:

  1. geographical territories

Eastern Scotland gets a lot of sun. (Paul 1994: 268) 

  1. political entities

Britain has declared war on San Marino. (Cruse 2000: 117)

  1. groups of people

Indianapolis voted for the referendum. (Nunberg 1995: 117)

and even seem to change the referent in the middle of a sentence:

Belgium is covered by a metre of snow, which is why it has asked France for help in clearing it. (Viebahn 2018: 754)
London is an ugly city but it is well run. (Borg 2012: 153)
London is so unhappy, ugly and polluted that it should be destroyed and rebuilt 100 miles away. (Chomsky 2000: 37)

As Viebahn (2018: 754) puts it for (12), “‘Belgium’ stands for a geographical region, but the first anaphoric ‘it’ for a political entity.” In such examples it is unclear which of the potential referents is the primary one and even if there is one primary.[6] In this paper I will only be concerned with the systematic reference of names of cities and countries to geographical territories, political entities and inhabitants – such as in (9)–(14) – which all seem to be equally plausible candidates for the default referents (cf. Pethö 2001). After considering and rejecting deferred reference as a tool for the analysis of such cases (Sections 2 and 3), I argue for treating the names of cities and countries as referring to a single complex referent, in analogy to the treatment of book or city in Asher and Pustejovsky’s dot type semantics.

2 Deferred reference

The most comprehensive analysis of deferred reference has been proposed by Geoffrey Nunberg (1993, 1995). Although Nunberg was skeptical about the deferred uses of proper names (cf. Nunberg 1992), it can be argued that there are uses of proper names which pass linguistic tests analogous to those deployed by Nunberg (1993, 1995) to argue for deferred reference of indexicals. In this section, I consider deferred reference as a tool for the analysis of some uses of proper names but in the following section I will argue against such an analysis of (9)–(14).

2.1 What is deferred reference?

Deferred reference occurs when we refer to one thing by way of another, where the latter is somehow given by the context of utterance. Nunberg called the auxiliary object given by context an index, reserving the term referent for the intended semantic value of the expression used. The index corresponds to the referent in a contextually salient manner. According to Nunberg, deferred reference is constrained by the meaning components of an expression, specifying the grammatical features which must be in agreement with either the index or the referent. A paradigmatic example is referring to the father of a girl with the masculine pronoun he, while pointing at the girl:

He is in real estate. (Nunberg 1993: 26)

or referring to an author while pointing at a book:

She was my chemistry teacher. (Nunberg 1993: 26)

In both cases the referent must agree in number, gender and animacy features with those encoded in the meaning of the pronoun used, while the index is given by demonstration (for demonstratives) or by the Kaplanian character (for other indexicals).

Nunberg (1995) used these requirements to advocate that (17) should be treated as a case of deferred reference (said while handing the keys to a car which is parked out back to a car park attendant):

This is parked out back. (Nunberg 1995: 110)

“[W]e would be inclined to say that the subject refers not to the key that the speaker is holding, but to the car that the key goes with. And in fact all the linguistic evidence supports this analysis. For example, the number of the demonstrative is determined by the intended referent, not the demonstratum. So even if the customer is holding up several keys that fit a single car, he would say ‘This is parked out back,’ whereas if he’s holding up a single key that fits several cars, he would say, ‘These are parked out back.’”(Nunberg 1995: 110)

In contrast, (18), uttered in the same situation, cannot be treated as deferred reference, due to the grammatical features of I, which require that the referent be animate:

I am parked out back. (Nunberg 1995: 110)

As Nunberg puts it, “the verb…is first-person, not third-person as we would expect if the subject referred to a car” (1995: 128, n. 3). For (18), Nunberg proposes an analysis in terms of meaning transfer involving the predicate, according to which “the predicate parked out back contributes a property of persons, the property they possess in virtue of the locations of their cars” (1995: 111), while the pronoun refers directly to the speaker (cf. also Kijania-Placek 2020; Kijania-Placek and Banaś 2021).

2.2 Deferred reference of proper names

Some uses of proper names seem to behave in a manner analogous to (17). In (5), repeated here:

Obama left the Halloween party very soon.

Barack Obama, being the default referent of the name Obama, can serve as the index, which points to the person dressed as Obama, corresponding to the index by way of being dressed as Obama. Proper names in English do not possess many grammatical features which would allow for testing, but plurality and anaphoric reference are such features. If the person dressed as Barack Obama were a woman, we could comment on her behavior by uttering (19):

She must have had one too many.

Assuming (19) was uttered immediately after (5), she is anaphoric on Obama and is feminine, which supports the hypothesis that Obama in (5) refers in a deferred way to the only woman dressed as Barack Obama at this particular party. If two people were dressed as Obama, we could appraise their costumes by saying (20):

The Obamas have the best costumes.

Again, the grammatical features of the names follow those of the referents, not those of the index. An example involving geographical names is (6), repeated here:

My parents protested during Vietnam.

Here the index is the country, but the deferred referent is the Vietnam War. This interpretation is supported by the grammatical requirements of the preposition during, which demands an eventive complement. Since a country is not an event, the referent is plausibly the most salient event that corresponds to it (cf. Kijania-Placek and Banaś 2021).

The working hypothesis here is that the index for proper names is given by the naming convention which is relied upon in the context and the referent corresponds to the index in a contextually salient manner. As always with deferred reference, the referent must agree with the name in whatever grammatical features the name possesses. Languages other than English with more grammatical features that are morphologically marked, such as Czech or Polish, are better suited for testing this hypothesis and crosslinguistic analysis gives it preliminary support. I will not pursue that analysis here, however, because although I believe that it can be sustained for examples such as (5), (6), (20) or (21) below (for an extended argument see Kijania-Placek and Banaś 2021), in the following section I will try to show that deferred reference is not sustainable for the analysis of examples such as (9)–(14), which are my main concern in this paper.

2.3 Deferred reference of geographical names

Although deferred reference might arguably work for geographical names used as in (6) or (21):

John was shot during Vietnam.[7] (Srinivasan and Rabagliati 2015: 134)

the same analysis would fail for (9)–(14). To see this, I will turn to a language with rich morphological marking of the grammatical features of proper names, one which allows number and gender agreement to be exploited in the grammatical tests. Nunberg used that strategy in his argument for the deferred reference of (17). By turning to Italian, where demonstratives and adjectives are marked for gender, he was able to rely on gender agreement: “the word for key is feminine, la chiave, and the word for truck is masculine, il camion. If a customer gives the attendant the key to a truck, it will be the referent, not the demonstratum, that determines the gender of the demonstrative and the adjective for ‘parked’” (Nunberg 1995: 110). Thus, when holding up a key (Italian: la chiave, fem., sg.) to refer to a truck (Italian: il camion, masc.), he would say:

‘This is parked in back.’ (Nunberg 1995: 110)

in compliance with the meaning requirements of the pronoun. An even more inflected language is Polish, where not just personal pronouns and adjectives but also demonstratives and verbs (in past tense) are sensitive to the grammatical features of proper names. Before turning to examples such as (9)–(14), I will demonstrate the testing strategy on an example that seems to work in an analogous way to (5) and (6):[8]

France played Germany tonight.

It seems that in (22) the country of France could be the index, and the team that represents the country would be the deferred referent. Since proper names are not marked for gender in English, it is not possible to establish deferred reference by manipulating this feature. Additionally, the nature of the correspondence between cities and their teams does not allow grammatical number to be manipulated in the manner adopted by Nunberg in (17) (see the quote in Section 2.1 above). But gender agreement can be tested in Polish, where the proper names of cities have been assigned gender. (22) translated into Polish is (22′):


If the country is the index and the team the deferred referent, the gender of the verb should agree with the grammatical gender of team – drużyna, which is feminine in Polish. This, however, turns out to be a coincidence – France (‘Francja’) is feminine as well – which can be shown by changing the order of the names in (22). If (22) is analyzable as deferred reference, so should be (23):

Germany played France tonight.
wieczór zFrancją.

In (23′) the verb agrees in number and gender with the name, but not with the potential referent which is the team (still feminine and singular), excluding thus the analysis of (22) or (23) as deferred reference. If (22) and (23) were a case of deferred reference, (23′) should be infelicitous. Not because it is ungrammatical, it is grammatical, but because the gender of the verb does not match the gender of the default referent.

I will now use the same line of argumentation to show that examples (9)–(14) cannot be analyzed by assuming, for example, the geographical territory to be the primary referent and treating the political entity and the inhabitants as deferred referents. If (11), repeated here:

Indianapolis voted for the referendum.

were a case of deferred reference, then in Polish the verb should be in gender and number agreement with the referent, but it is not, as can be seen by considering names differing in gender and number in the same construction (cf. also (Kijania-Placek and Banaś 2021):

Barcelona voted for the referendum.
Katowice voted for the referendum.

In all (11), (11′), (24), (24′), (25) and (25′) the intended referents are the inhabitants of the respective cities. But whatever the number and gender features associated with the referents,[10] the features should remain constant in (11′), (24′) and (25′), because the type of the referent does not change. Instead, the agreement pattern follows the number and gender of the proper names themselves, in violation of the constraints of deferred reference. Analogous examples can be formulated to show that (10):

Britain has declared war on San Marino.

cannot be analyzed as deferred reference either.[11] In fact, since the arguments were independent of what was taken as the index, I have shown that neither the geographic territory, nor the political entity or the inhabitants, which are intuitive referents in examples (9)–(14), can be interpreted as deferred referents. As we can see, in contrast to the deferred reference interpretation that requires grammatical features to follow those of the referents, in Polish the grammar follows the gender and plurality characteristics of the proper name in all cases – for example, mieszkańcy (‘inhabitants’) are masculine personal in Polish. This is particularly discernible in examples such as (14) and its Polish counterpart (14′), which combine two types of potential referents, or (26) and (26′), which combine all three:

London is so unhappy, ugly and polluted that it should be destroyed and rebuilt 100 miles away.
iwybudowanyna nowo100 mildalej.
andbuild.PAST.3SG.MSCanew100 milesaway
Britain, despite the fact that it is lying under one metre of snow and is mourning the death of the Queen Mother’s corgi, has declared war on San Marino. (Cruse 2000: 117)
Wielka Brytnia,mimożejestpokrytametrem
Britain.SG.FEMdespitethatiscover.PAST.3SG.FEMone-meter snow
śnieguijestwżałobiepocorgiKrólowej Matki,
snowandisinmourningaftercorgiQueen Mother’s
wypowiedziaławojnęSan Marino.
declare.PAST.3SG.FEMwarSan Marino.DAT

In (14) unhappy concerns the inhabitants, while ugly, polluted, destroyed and rebuilt – the territory or physical aspects of the city. Although territory (‘terytorium’) in Polish is neuter, and inhabitants (‘mieszkańcy’) masculine personal, all gender agreements follow the masculine gender of the proper name.[12] Similarly with (26), where covered concerns a geographical territory, in mourning concerns the population and declare a political entity, all expressions are in gender agreement with the gender of the proper name, disregarding the varying genders of the potential deferred referents.[13] From the grammatical point of view, it seems that Britain is just a territory, a political entity, and its inhabitants/citizens taken as one complex entity, whatever its “real” metaphysical status. Different predicates select particular aspects of the object, the way predication selects for particular aspects in the case of nouns such as book or city. In what follows, I propose an analysis of examples such as (9)–(14) based on the treatment of book and city by Asher and Pustejovsky.

3 The semantic properties of book and city

Interpretation processes such as deferred reference rely on one sense of a word being the default one and subject to meaning or reference transfers. Yet with many words there does not seem to be a clear sense of which of its many senses is the default one. As Nunberg put it: “there is a substantial class of cases where we have no principled grounds for deciding which of several uses is conventional” (Nunberg 1979: 154; cf. also Carston 2012). A canonical example is the noun book, which has at least two candidates for the primary sense and the noun can be used to talk about any of them (Bosch 2007; Carston 2012; Chomsky 2000; Pustejovsky 1995, 2005):

  1. a physical object

Mary burned my book on Mahler. (Pustejovsky 2005: 6)

  1. a set of ideas, content, information

Mary believes all of Chomsky’s books. (Pustejovsky 2005: 6)

These senses (content vs. physical object) can be combined in one sentence, as in Chomsky’s famous examples:

The book that he is planning will weigh at least five pounds if he ever writes it. (Chomsky 2000: 180)
His book is in every store in the country. (Chomsky 2000: 180)

Or Carston’s:[14]

Peter is reading the novel he found at the bus-stop. (Carston 2012: 616)

Even the first part of (31):

Peter is reading the novel.

deploys both the physical and informational aspects of book, because, as Pustejovsky and Chomsky insisted, the verb read requires a complement that is both informational and physical at the same time. This last feature is especially important, as it shows that the two aspects are not disjunctive and that the referent of book is the “whole” book, with both of its aspects (cf. Asher 2011). This behavior is in contrast with homonymous words, which behave differently in the contradiction test, where a sentence featuring two occurrences of the expression and one occurrence of negation can be true, as it is for bank but not for book (cf. Viebahn 2018):

This is a bank, but it is not a bank [in a different sense]. (Viebahn 2018: 758)
*That is a book, but it is not a book. (Viebahn 2018: 758)

This shows that when we point at a physical book, we point at an informational book at the same time. Similar ideas are proposed by Cruse (he uses the term facet for what Asher and Pustejovsky call aspect, see below; cf. also Asher 2011; Asher and Pustejovsky 2005; Cruse 1986; Dölling 2020):

“Facets can be described as fully discrete but non-antagonistic readings of a word. Another important characteristic is that they are characteristically of distinct ontological types. However, in spite of their discreteness and ontological distinctness it would not be correct to say that they represented distinct concepts: they are somehow fused into a single conceptual unit. Amongst the evidence for this claim the following may be cited:

  1. Ordinary speakers are not normally aware of the dual nature of book: it has to be pointed out to them (however, once pointed out, it becomes obvious). The facets form a single, unified ‘gestalt’. The default reading of book is the combined one.

  2. As we have seen, predicates selecting different facets can co-ordinate without zeugma, and there is no normal requirement for speakers to intend, or hearers to identify, only one of the facets, as is the case with true ambiguity.

  3. The combined reading functions as a basic level item.

  4. Some predicates require both facets to be present: publish a book, John is reading a book.

  5. The combined meaning can be metaphorically extended: John’s mind is a closed book to me. This cannot be construed unless one takes into account both facets.

  6. The combination may bear a proper name (e.g. Britain).” (Cruse 2000: 116)

Aspects are not physical parts of an object. As Asher insisted: “[a]ll of the book (in its material aspect) is a material object, and all of the book (in its informational aspect) is also an informational object. And so we can’t conclude that the book isn’t a material object (or an informational object) at some time, state or location” (2006: 14; see also Collins 2017). It is important to note, however, that the aspects of book also cannot be equated with the type/token distinction, as can be seen in the interplay between the noun and the counting principles (Asher and Pustejovsky 2005):

The student read every book in the library. (Asher and Pustejovsky 2005: 11)
The student carried off every book in the library. (Asher and Pustejovsky 2005: 11)
John has stolen every book there is. (Asher and Pustejovsky 2005: 14)

While in (35) what is counted are the informationally distinguished books, in (36) the quantifier is only sensitive to the physical object aspect of book. Yet the default interpretation of (37) requires John stealing one (physical) copy of every book in the informational sense. Additionally, the physical aspects of book alone already require a counterpart of the type-token distinction, which can be seen in the distinction between a particular copy of a book (38), the physical properties of an edition (39) and a purely informational aspect of a book (40) (Copestake and Briscoe 1995):

That book is covered with coffee. [physical object – token] (Copestake and Briscoe 1995: 53)
That book has an unreadable font. [physical object – type] (Copestake and Briscoe 1995: 53)
That book is full of metaphorical language. [content] (Copestake and Briscoe 1995: 53)

Thus far I have concentrated on the properties of book, but the phenomenon of many-aspect words is ubiquitous in natural language. For example, dictionary, newspaper, map, letter, film or CD all have physical and informational aspects; speech, lecture, movie or play have eventive and informational aspects; nouns such as bank, school, university, parliament, church or opera designate institutions and the buildings those institutions reside in; lunch, dinner, as well as human and city (see below) are reported to “refer to entities which in each case seem to have a dual nature” (Dölling 2020: 9; see also Asher and Pustejovsky 2000; Ortega-Andrés and Vicente 2019; Pustejovsky 2005). As (38)–(40) show, duality – physical object vs. content – does not really do justice to the semantic complexity of book. The situation is even more obvious with newspaper, which has at least three basic aspects: copy/physical object (41), information/content (42), and organization/institution/publisher (43) (based on examples in Pustejovsky 2005):

The cup is on top of the newspaper.
I disagreed with the newspaper.
The newspaper fired its editor.

They can combine in copredication and anaphoric references:[15]

She handed me an envelope containing all the newspapers she had worked for. (Nunberg 1995: 126) [physical object vs. organization]
The newspaper is full of coffee stains and turned out to be very uninteresting. (Dölling 2020: 12) [physical object vs. content]
Mary was fired by the newspaper and so she vowed to never read it again.
(Antunes and Chaves 2003: 7) [organization vs. content]

For our present purposes, an important many-aspect word is city (Asher and Pustejovsky 2000; Dölling 2020; Pustejovsky 1998). It exhibits at least the location/territory aspect (47), the organization/political entity aspect (48) and the inhabitants/collection of people aspect (49) (Asher and Pustejovsky 2000):

Many cities have good restaurants. (Asher and Pustejovsky 2000: 9)
Many cities have adopted anti-smoking legislation. (Asher and Pustejovsky 2000: 9)
Many cities vote Democratic. (Asher and Pustejovsky 2000: 9)

They all combine felicitously, as in the following examples:

The city that John lives in just passed a law against smoking in bars.
(Asher and Pustejovsky 2000: 8) [territory vs. organization]
The city is extremely polluted, but voted against smoking in public places last year. [territory vs. people]
Most cities that vote democratic passed anti-smoking legislation last year.
(Asher 2011: 132)[territory vs. people]

Pustejovsky (1995) suggested an influential semantics for the analysis of many aspect words, which was later developed in a serious of works co-authored with Nicolas Asher and by Asher himself. In the following Section, I will briefly introduce Pustejovsky’s proposal and in Section 5 I will deploy the theory to the analysis of the proper names of cities and countries.

4 •-type words

Pustejovsky’s ideas were motivated by the phenomenon of copredication, i.e. the grammatical construction in which two predicates apply to the same argument.[16] Predicates select for specific argument types but also some arguments select the predicates that govern them (Pustejovsky 1991). In general, predication is only successful if the predicate and its arguments are of compatible types, as in (53), where the predicate heavy requires an argument to be a physical object:

This suitcase is heavy.

Copredication is unproblematic when the argument fulfills the requirement of both of the predicates (54) but typically renders an expression infelicitous if it is not so (55):

He vandalized and then burned down the shop.
*Cats are mammals and sell well in toy shops.

In some cases involving nouns such as book, dinner, city or those mentioned in the previous section, however, copredication is successful even though the requirements of the predicates are conflicting. Thus (56) is perfectly felicitous, even though burn requires a physical object while for on Mahler an informational one is needed (Pustejovsky 2005):

Mary burned my book on Mahler.

A similar phenomenon occurs in anaphoric reference, as exemplified by (56):

John’s mother burned the book on magic before he mastered it. (Asher 2011: 170)

The first predicate requires the argument be a physical object, while the pronoun it refers anaphorically to an informational one – the content of the book. Pustejovsky (1995) categorizes all nouns by types, but argues that copredication phenomena supports complex typing for words such as book. What he calls dot type or •-type is formed by two or more simple types (simple types are physical object, informational object etc.) but are not their ordinary sums. While some predicates require one of the simple types, as exemplified by (27), (54), and (55), other require an argument of the complex type. The example Pustejovsky mentions in connection with book are the verbs read and write, which require an object which is physical and informational at the same time.[17] Constituent types also support divergent counting principles for the •-type words, which explains the different truth conditions for (35) and (37) above.

According to Pustejovsky and Asher (Asher and Pustejovsky 2005; Pustejovsky 1995) what explains copredication is that for •-type words the constituent types correspond to aspects of objects denoted by those words and those aspects are available simultaneously – at least for words with two constituent types – during composition of the meaning of the whole sentence. The authors alternatively use the term dot object in connection with •-type words but insists that dot object “refers to objects with a complex type (not to complex objects – whatever those might be – or to pairs of objects), with several aspects which have become part of the meanings of the words that denote such objects. Such dot objects allow for predications which are licensed over either of the two dot element types” (Asher and Pustejovsky 2005: 7).[18] • is a binary type construction operator but it can be iterated, as is required by newspaper, whose type consist of three simple types and is structured as organization•(content•physical object) (Pustejovsky 2005).

Asher and Pustejovsky propose to treat all words mentioned in Section 3 as •-types but those relevant for our analysis of proper names are the words city and country. As for newspaper, the •-types of city and country consist of (at least) three types, corresponding to the territory, organization/government and inhabitants/citizens aspects of cities and countries, which, as newspaper, are structured as double •-types: people•(territory•political entity) (Asher and Pustejovsky 2000; Pustejovsky 1998, 2005). The constituent aspects are available for predication, as exemplified by (47)–(52) above, as well as by (57)–(59):

John travelled to the country last month. (Dölling 2020: 4)[territory]
So, do you think a country has the right to kill an unarmed man?
(Galery 2012: 46)[political entity]
The country voted against the EU constitution in the 2005 referendum.[19]

In what follows I will argue that proper names of cities and countries inherit the semantics of the respective common nouns and should be analyzed as •-types.

5 Names of cities behave like city

Although the theses of type theorists primarily concerned common nouns, the examples they use include the proper names of both countries and cities. Cruse’s analysis did not involve the semantic machinery of types, but his concept of facets is akin to that of aspects and he explicitly suggested that some proper names may be considered as multifaceted words (Cruse 2000, cited above). The thesis I propose here is that the proper names of cities and countries are such multi-aspect words and that they should be analyzed as Pustejovsky’s •-type words, i.e. as a people•(territory•political entity) types.[20] The aspect labeled political entity concerns the abstract, institutional features of a city/country, while those labeled territory concern its physical aspects and people its human ones. As is the case for book and city, predication can select specific constituent aspects (some examples are repeated from previous sections):

Britain lies under one metre of snow.[territory]
Eastern Scotland gets a lot of sun.[territory]
Brazil is the largest country in South America.
(Ortega-Andrés and Vicente 2019: 1)[territory]
Liverpool is spread out. (Retoré 2014: 214)[territory]
Indianapolis voted for the referendum.[people]
Britain mourns the death of the Queen Mother’s corgi.
(Cruse 2000: 117)[people]
The south side of Cambridge voted Conservative.
(Copestake and Briscoe 1995: 27)[people]
France voted against the EU constitution in the 2005 referendum.
(Evans 2015: 106)[people]
France is one of the leading nations in the European Union.
(Evans 2015: 106)[people]
Britain has declared war on San Marino.[political entity]
London said that a new passport could not be issued.
(Copestake and Briscoe 1995: 43)[political entity]
Brazil is a republic. (Ortega-Andrés and Vicente 2019: 1)
[political entity]

In the examples above, predication concerns different aspects of the respective cities and countries but none of the aspects seems to be the primary one. They can easily combine in copredication and anaphora:

Germany is a central European democracy. (Collins 2017: 680)
[territory vs. political entity]
Belgium is covered by a metre of snow, which is why it has asked France for help in clearing it. [territory vs. political entity]
London is an ugly city but it is well run.[territory vs. political entity]
London is so unhappy, ugly and polluted that it should be destroyed and rebuilt 100 miles away. [people vs. territory]
London tends to vote Conservative, despite being the largest urban area in the U.K. (Collins 2017: 680) [people vs. territory]
Britain, despite the fact that it is lying under one metre of snow and is mourning the death of the Queen Mother’s corgi, has declared war on San Marino. [territory vs. people vs. political entity]
Brazil is a large Portuguese-speaking republic that is very high in inequality Rankings.[21] [territory vs. people vs. political entity]
Brazil is a large two-century-old Portuguese-speaking country.
(Arapinis and Vieu 2015: 290)[territory vs. political entity vs. people]

I propose that in the kind of uses exemplified by (9)–(14), in contrast to the predicative, deferred and descriptive uses exemplified by (4), (6) or (8) respectively, the proper names of countries and cities directly refer to their objects of reference, regardless of whether the focus is on the territorial (physical), political or human aspects of the referents. While lexical semantics is informed by common sense metaphysics, which provides the basic types for words and the basic relations between types, the thesis advanced here should not be taken as one about the metaphysics of the objects referred to by the names Paris or Brazil. The claim would rather be that, regardless of the metaphysical status of cities, Paris itself would inherit it.

Part of the thesis I advance is that the kinds of aspects that enter the lexical entry underlying the direct use of a proper name are those common to all names of a particular category, i.e. those intrinsic to its being a city or country respectively. Thus, while I am open to including additional aspects on the list, I oppose many authors directly or indirectly relating their theories to •-types, who include features specific only to some cities among constituent types (or their equivalents thereof). An important, often included aspect is representation as in that by a sports team. Thus, e.g. in Ortega-Andrés and Vicente (2019), Retoré (2014), or Evans (2015) we encounter examples such as (73), (75) or (76) indiscriminately mixed with examples of the territory, people or political entity type (some examples are repeated from above):

Liverpool is spread out.[territory]
Liverpool voted. (Retoré 2014: 214)[people] or [political entity]
Liverpool won. (Retoré 2014: 214)[representation]
France is a country of outstanding natural beauty. (Evans 2015: 106)
France is one of the leading nations in the European Union.[people]
France beat New Zealand in the 2007 Rugby world cup. (Evans 2015: 106)
France voted against the EU constitution in the 2005 referendum.
Brazil is a republic.[political entity]
Brazil has won five World Championships.
(Ortega-Andrés and Vicente 2019: 1)[representation]
Brazil is the largest country in South America.[territory]

To argue for my exclusive position, I will start by making a methodological point. The distinction I propose, between aspects which are included in the •-type of a name and those that are excluded, is a principled one which is based on the inherent, necessary features of cities and countries respectively, whatever they are. The alternative proposal is arbitrary and lucking a criterion for selecting the aspects. Ortega-Andrés and Vicente (2019) are, as far as I know, the only authors who may be taken to have indirectly argued for including sports representation among constituent types, but at the same time they opted against including the exam aspect for school on the basis of exams not being necessary for the school’s existence. Yet, since not all cities have sports teams which represent them, that aspect is equally unnecessary. Still, we need a criterion, because representation is not the only potential aspect that the proper names of cities and countries are used to depict. One more candidate for an aspect could be an event connected with a city or country, as in (6), repeated from Section 1:

My parents protested during Vietnam.

In Section 2.2 I proposed an analysis of this example by deferred reference but there seems to be no reason why Ortega-Andrés and Vicente, Retoré, or Evans should exclude the event aspect from •-type (or equivalent) analysis of Vietnam and other names of cities. Other aspects could be considered and thus a criterion is required on pain of a theory lacking explanatory power.

My next argument comes from an intuitive difference between the basic territorial, political and human aspects of cities and countries on the one hand and other aspects, such as representation or event. While choosing a primary referent among territory, people or political entity might be a challenge and all are equally available for selection in predication, it seems clear that neither the Vietnam war nor the sports team are the primary referents of the respective proper names. Rather, as in deferred reference, the interpretation process proceeds through the country of Vietnam to the war, which is not an internal aspect of the country but corresponds to it in a contextually salient manner.[22]

Finally, I turn to linguistic data to support my thesis. While there are some cases of successful co-predication (see Ortega-Andrés and Vicente 2019), in most cases co-predication between the basic aspects and the representation aspect fail, as some of the authors themselves notice:

Liverpool is spread out and voted (last Sunday). (Retoré 2014: 214)
[territory vs. people]
*Liverpool voted and won (last Sunday). (Retoré 2014: 214)
[territory vs. representation]
*Liverpool won the cup and voted against having a mayor.
(Retoré 2013: 5)[representation vs. people]
*Barcelona plays Real Madrid tonight and is the fifth most populous urban area in the European Union. [representation vs. territory]

This puts the sports representation aspect apart from territory, people and political entity aspects, which are easily available for both co-predication and anaphora.

6 Conclusion

In this paper I have proposed a •-typed based interpretation of some uses of proper names of cities and countries. I have argued that an aspect of a city (or country) enters the semantics of its name only if it is constitutive for its being a city (or country). The aspects distinguished are thus those labeled territory, people and political entity. I have argued against including sports representation of a country or city or an event corresponding to it among aspects included in the •-type of a proper name.

The complex, many aspect nature of the referents of names of cities and countries can be traced back to them being social artefacts. A natural extension of the proposal would thus be a semantics of other institutional names, including the names of schools, universities or legal proper names such as the Constitution of the United States or The European Parliament. This requires further research.

Since none of the distinguished aspects is the basic one for a proper name like Paris, the name refers directly to an object which somehow encompasses all of these aspects. However, the proper names of countries and cities, as well as personal proper names, additionally have predicative, deferred, or descriptive uses, which are based on semantic mechanisms other than direct reference. I share an inclusive approach to semantics that aims to incorporate all systematic uses of expressions of a type in the analysis of its meaning (Levinson 2000). In the case of proper names, such an approach naturally leads to proposing a polysemous account of proper names.[23] That, however, is a topic for another paper.

Corresponding author: Katarzyna Kijania-Placek, Institute of Philosophy and the Jagiellonian Center for Law, Language, and Philosophy, Jagiellonian University, Krakow, Poland, E-mail:

Funding source: Narodowe Centrum Nauki

Award Identifier / Grant number: UMO-2015/19/B/HS1/01068

Funding source: Narodowa Agencja Wymiany Akademickiej

Award Identifier / Grant number: PPI/APM/2018/1/00022


I would like to thank Olga Poller and Paweł Banaś for discussing some of the examples concerning deferred reference. Section 2.1 and 2.2 as well as the discussion of examples (6), (11), (24), and (25) are based on a paper co-authored with Paweł Banaś (Kijania-Placek and Banaś 2021). This work has been supported by the (Polish) National Science Centre (grant no. UMO-2015/19/B/HS1/01068) and by the NAWA grant no. PPI/APM/2018/1/00022.

  1. Dedication: This paper is dedicated to the memory of Jerzy Pelc, whose pioneering work on contextual involvement in the interpretation of natural language expressions influenced generations of researchers. The present author has been especially influenced by the ideas developed in his book O użyciu wyrażeń (1971).


Antunes, Sandra & Rui P. Chaves. 2003. On the licensing conditions of co-predication. In P. Bouillon & K. Kanzaki (eds.), Proceedings of the second international workshop on generative approaches to the lexicon, 161–168. Geneva.Search in Google Scholar

Arapinis, Alexandra & Laure Vieu. 2015. A plea for complex categories in ontologies. Applied Ontology 10. 285–296. in Google Scholar

Asher, Nicholas. 2006. Things and their aspects. Philosophical Issues 16. 1–23. in Google Scholar

Asher, Nicholas. 2011. Lexical meaning in context: A web of words. Cambridge: Cambridge University Press.10.1017/CBO9780511793936Search in Google Scholar

Asher, Nicholas & James Pustejovsky. 2000. The metaphysics of words in context. (accessed 31 August 2020).Search in Google Scholar

Asher, Nicholas & James Pustejovsky. 2005. Word meaning and commonsense metaphysics. (accessed 31 August 2020).Search in Google Scholar

Bach, Kent. 2002. Giorgione was so-called because of his name. Philosophical Perspectives 16. 73–103. in Google Scholar

Bach, Kent. 2015. The predicate view of proper names. Philosophy Compass 10. 772–784. in Google Scholar

Boër, Steven E. 1975. Proper names as predicates. Philosophical Studies 27. 389–400.10.1007/BF01236458Search in Google Scholar

Borg, Emma. 2012. Pursuing meaning. Oxford: Oxford University Press.10.1093/acprof:oso/9780199588374.001.0001Search in Google Scholar

Bosch, Peter. 2007. Productivity, polysemy, and predicate indexicality. In H. Zeevat & B. ten Cate (eds.), Proceedings of the sixth international Tbilisi symposium on language, logic and computation, 58–71. Dordrecht: Springer.10.1007/978-3-540-75144-1_5Search in Google Scholar

Burge, Tyler. 1973. Reference and proper names. Journal of Philosophy 70. 425–439. in Google Scholar

Carston, Robyn. 2012. Word meaning and concept expressed. Linguistic Review 29. 607–623. in Google Scholar

Chatzikyriakidis, Stergios & Zhaohui Luo. 2015. Individuation criteria, dot-types and copredication: A view from modern type theories. In Proceedings of the 14th international conference on mathematics of language, 39–50. Chicago, IL: ACL.10.3115/v1/W15-2304Search in Google Scholar

Chomsky, Noam. 2000. New horizons in the study of language and mind. Cambridge: Cambridge University Press.10.1017/CBO9780511811937Search in Google Scholar

Collins, John. 2017. The copredication argument. Inquiry 60. 675–702. in Google Scholar

Copestake, Ann & Ted Briscoe. 1995. Semi-productive polysemy and sense extension. Journal of Semantics 12. 15–67. in Google Scholar

Cruse, D. Alan. 1986. Lexical semantics. Cambridge: Cambridge University Press.Search in Google Scholar

Cruse, D. Alan. 2000. Meaning in language: An introduction to semantics and pragmatics. Oxford: Oxford University Press.Search in Google Scholar

Dölling, Johannes. 2020. Systematic polysemy. In Daniel Gutzmann, Lisa Matthewson, Cecile Meier, Hotze Rullmann & Thomas E. Zimmermann (eds.), The Wiley Blackwell companion to semantics, 1–27. Oxford: Wiley Blackwell.10.1002/9781118788516.sem099Search in Google Scholar

Evans, Vyvyan. 2015. A unified account of polysemy within LCCM theory. Lingua 157. 100–123. in Google Scholar

Falkum, Ingrid Lossius & Agustin Vicente. 2015. Polysemy: Current perspectives and approaches. Lingua 100. 1–16. in Google Scholar

Frisson, Steven. 2009. Semantic underspecification in language processing. Language and Linguistics Compass 3. 111–127. in Google Scholar

Galery, Thiago N. 2012. Descriptive pronouns revisited. The semantics and pragmatics of identification based descriptive pronominal interpretations. University College London. PhD diss.Search in Google Scholar

Jeshion, Robin. 2015. Referentialism and predicativism about proper names. Erkenntnis 80. 363–404. in Google Scholar

Kijania-Placek, Katarzyna. 2018. Can minimalism about truth embrace polysemy? Synthese 195. 955–985. in Google Scholar

Kijania-Placek, Katarzyna. 2020. Descriptive indexicals, deferred reference, and anaphora. Studies in Logic, Grammar and Rhetoric 62. 25–52.10.2478/slgr-2020-0012Search in Google Scholar

Kijania-Placek, Katarzyna & Paweł Banaś. 2021. Deferred reference of proper names. Journal of Semantics. in Google Scholar

Leckie, Gail. 2013. The double life of names. Philosophical Studies 165. 1139–1160.10.1007/s11098-012-0008-3Search in Google Scholar

Levinson, Stephen C. 2000. Presumptive meanings: The theory of generalized conversational implicature. Cambridge, MA: MIT Press.10.7551/mitpress/5526.001.0001Search in Google Scholar

Matushansky, Ora. 2006. Why rose is the rose: On the use of definite articles in proper names. Empirical Issues in Syntax and Semantics 6. 285–307.Search in Google Scholar

McElree, Brian, Steven Frisson & Martin J. Pickering. 2006. Deferred interpretations: Why starting Dickens is taxing but reading Dickens isn’t. Cognitive Science 30. 181–192. in Google Scholar

Nunberg, Geoffrey. 1979. The non–uniqueness of semantic solutions: Polysemy. Linguistics and Philosophy 3. 143–184. in Google Scholar

Nunberg, Geoffrey. 1992. Two kinds of indexicality. Semantics and Linguistic Theory 2. 283–302. in Google Scholar

Nunberg, Geoffrey. 1993. Indexicality and deixis. Linguistics and Philosophy 16. 1–43. in Google Scholar

Nunberg, Geoffrey. 1995. Transfers of meaning. Journal of Semantics 12. 109–132. in Google Scholar

Ortega-Andrés, Marina & Agustín Vicente. 2019. Polysemy and co-predication. Glossa 4. in Google Scholar

Paul, Matthias. 1994. Young Mozart and the joking Woody Allen: Proper names, individuals and parts. Semantics and Linguistic Theory 4. 268–281. in Google Scholar

Payne, John & Rodney Huddleston. 2002. Nouns and noun phrases. In R. Huddleston & G. K. Pullum (eds.), The Cambridge grammar of the English language, 323–523. Cambridge: Cambridge University Press.10.1017/9781316423530.006Search in Google Scholar

Pelc, Jerzy. 1971. O użyciu wyrażeń [On the use of expressions]. Wrocław: Ossolineum.Search in Google Scholar

Pethö, Gergely. 2001. What is polysemy? A survey of current research and results. In E. Nemeth & K. Bibok (eds.), Pragmatics and flexibility of word meaning, 175–224. Amsterdam: Elsevier.10.1163/9780585474267_011Search in Google Scholar

Pustejovsky, James. 1991. The generative lexicon. Computational Linguistics 17. 409–441.Search in Google Scholar

Pustejovsky, James. 1995. The generative lexicon. Cambridge, MA: MIT Press.Search in Google Scholar

Pustejovsky, James. 1998. The semantics of lexical underspecification. Folia Linguistica 32. 323–347. in Google Scholar

Pustejovsky, James. 2005. A survey of dot objects. (accessed 31 August 2020).Search in Google Scholar

Retoré, Christian. 2013. COmplexite et LANgage une etude formelle et experimentale des mecanismes de comprehension.∼retore/PRESENTATIONS/retore_slam_dec_2013.pdf (accessed 31 August 2020).Search in Google Scholar

Retoré, Christian. 2014. The montagovian generative lexicon Λt yn: A type theoretical framework for natural language semantics. In Ralph, Mathes & Aleksy, Schuber (eds.), Nineteenth international conference on types for proofs and programs (TYPES 2013), vol. 26, 202–229. Saarbrücken: Dagstuhl.Search in Google Scholar

Srinivasan, Mahesh & Hugh Rabagliati. 2015. How concepts and conventions structure the lexicon: Crosslinguistic evidence from polysemy. Lingua 157. 124–152. in Google Scholar

Tayebi, Sajed. 2018. In defense of the unification argument for predicativism. Linguistics and Philosophy 41. 557–576. in Google Scholar

Ueda, Tomoo. 2015. Analysis of geographical proper names in terms of the indexicality account of proper names. In Henning Christiansen, Isidora Stojanovic & George Papadopoulos (eds.), Modeling and using context, 316–327. Dordrecht: Springer.10.1007/978-3-319-25591-0_23Search in Google Scholar

Viebahn, Emanuel. 2018. Ambiguity and zeugma. Pacific Philosophical Quarterly 99. 749–762. in Google Scholar

Received: 2021-01-26
Accepted: 2021-01-31
Published Online: 2021-03-05
Published in Print: 2021-05-26

© 2021 Katarzyna Kijania-Placek, published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.