In the first type of binominal construction, N2 is a proper noun referring to a person. If the construction as a whole refers to a person, N1 can be a rank in a hierarchical organization like the army or the church, a title or profession, a form of address, or a kinship noun (especially tanteaunt and oomuncle). Some examples, mainly adapted from Haeseryn et al. (1997), are given in (152).

The resulting structures in (152) function as complex proper nouns, which is clear from the fact that they are usually bare, i.e. they cannot be preceded by an article. The examples in (153) show that they differ crucially in this respect from constructions in which the N1s occur by themselves.

It is also clear, as shown in (154), that most of these binominal constructions, like proper nouns, can be used both as vocatives and as regular arguments. Note that the title doctor in the (a)-examples can easily be replaced by the profession noun dokterdoctor; despite the difference in spelling, the two nouns have the same pronunciation, so in speech the difference in interpretation depends on the context.

Binominal constructions such as tante Jeanneaunt Jeanne should be distinguished from phrases such as mijn zuster Elsmy sister Els (i.e. with a determiner). This is clear from the fact that the latter cannot be used as a vocative; cf. the contrast between the two (b)-examples in (155). The proper noun Els can function as an appositive, which is clear from the distinctive intonation pattern in (155c'), with an intonation break before and after it; (155a') does not have this intonation pattern, but can perhaps be seen as a non-restrictive counterpart of (155c'). The unacceptability of (155c) shows that the proper noun Jeanne in tante Jeanne cannot be used as an appositive. For a discussion of appositions, see Section 3.1.3.

The examples in (156) show that most N1’s in (152) have the property that they can also be used as bare nominal predicates, although this is less common for meneer/mevrouwMr/Mrs in (152c), but they do occur in this function as the first member of binominal constructions: cf. Ik ben meneer VerdonkI am Mr. Verdonk.

Forms of address like meneer and mevrouw are also special in that they may be followed by a noun phrase denoting a high-status occupation or social function, as in (157a). When the second noun phrase denotes a less prestigious occupation or implies some subjective qualification, as in (157b), the complex noun phrase takes on an ironic connotation. The projection of N2 in cases such as (157) necessarily contains the definite article. N1, on the other hand, is never preceded by a definite article, again suggesting that the construction as a whole functions as a proper noun.

Constructions such as (157) also differ from those in (152) in that their use is more restricted. Their unmarked use is that of a vocative, and they can only be used in argument position when the person referred to is physically present. Thus, while (154b) can be uttered in the absence of the intended person, example (158b) seems to require that the intended person be physically present.

While the person nouns (N1’s) in the binominal constructions in (152) can typically also occur as bare nominal predicates, the reverse is not true: person nouns that can be used as bare nominal predicates cannot automatically be used as N1’s in constructions of the kind in (152). This is illustrated in (159) for the profession nouns meesterprimary school teacher and leraarsecondary school teacher; example (159a) shows that these nouns can both be used as nominal predicates, while (159b&c) show that only meester can easily be used as N1 in a binominal construction functioning as vocative or argument (not commonly done in English).

Schermer (2023) argues that the fact that the N1’s in (152) can also be used as bare nominal predicates attributing a property to their logical subjects indicates that the modification relation in such binominal constructions contradicts our earlier suggestion, based on the examples in (151), that the most common modification relation is the one where N2 has a modifying function with respect to N1. Instead, it seems that N1 functions as a modifier: tante Jeanne does not express that Jeanne is the aunt intended by the speaker, but that Jeanne is considered in her role as an aunt.

The N1’s in (152) are typically anumeric, i.e. they cannot be pluralized. This may be related to the fact that they are not used as referential expressions but as modifiers; like bare predicate nouns, they are property-denoting expressions lacking number features; cf. Mattens ( 1970 ) and Schermer (2023). This does not mean that the nouns cannot be pluralized (De Belder 2009), but if they are they require the presence of a determiner, as shown in (160a). We conclude that we are not dealing with a construction of the type in (152), but with a binominal construction of the type discussed in Subsection II; this is supported by the fact, illustrated in (160b), that such plural noun phrases cannot be used as vocatives.

If the construction as a whole refers to a geographical entity, N1 can be a noun denoting the set of geographical entities of which the referent of the whole binominal construction is a member. Some representative examples are given in (161). In such examples, the modification relation is often bidirectional: while it is clear that the proper name enables the hearer to identify the intended river, state, or city, it is also expressed that the proper name refers to a river, state, or city. Whether both directions are actually activated also depends on the hearer’s extra-linguistic knowledge: in (161b), it will be obvious to those speakers who are aware of the fact that the name Utrecht is used for both the province of Utrecht and its capital city. Note in passing that N2 can also be preceded by a definite article, provided that it also has one when the proper name is used in isolation (which may vary from case to case and speaker to speaker).

The binominal constructions in (161) differ from those discussed in Subsection I in that their N1’s are preceded by a definite article. This means that these N1’s are regular referential expressions, not anumeric property-denoting expressions lacking number features. This can also be seen from the fact, illustrated in (162), that they can be pluralize; cf. Schermer (2014/2023).

The bidirectional relation also seems to hold for examples such as (163). This may not be so clear in (163a), where it is obviously the proper name that modifies the noun familie and not vice versa, but it is in (163b), where it is simultaneously expressed that we are dealing with a writer named Jan Wolkers and that Jan Wolkers is a writer.

However, the semantic relation between the two nouns can also be asymmetrical: example (164a) is non-restrictive because the nominal part beroemde schrijver provides additional information about Jan Wolkers by saying that he is famous as a writer, whereas example (164b) is restrictive by expressing that Jan Wolkers is very famous as a writer but not as a sculptor; cf. Van de Velde (2009:78).

In the first edition of SoD, we reported that the definite article is obligatory in examples such as (163) and (164), but we missed that it was reported in Van der Horst & Van der Horst (1999:222ff.) that there is a current development of omitting the article; the authors report that this is especially common in sports coverage, but also regularly found elsewhere. The following two examples are taken from a larger collection of citations from the quality newspaper NRC Handelsblad (2013) in Schermer (2014).

The (optional) omission of the determiner leads to a construction that seems very similar to the binominal constructions discussed in Subsection I, but differs crucially from them in that it cannot be used as a vocative. The constructions with and without determiner also seem to differ in referentiality: in (166) the proper names (N2) are used to identify the intended referent of N1, whereas in (165) the N1’s are rather property-denoting expressions modifying N2.

Constructions such as (165) and (166) also differ in the attributive inflection of the adjectival modifier of N1; the inflection is obligatorily absent in determinerless constructions such as (165), as is also the case with bare nominal predicates, but present in constructions with a determiner such as (166); cf. Schermer (2023). Note that the attributive inflection in (166b) is optional; both cases can be found on the internet, which may be related to the fact that such A+P combinations can be more or less lexicalized or idomatic; cf. Section A27.1.2, sub II, for discussion.

Let us return to constructions with a determiner such as de schrijver Jan Wolkers. The examples in (167a) show that the linear order of the common and proper noun can sometimes be reversed, as in (167a). This example is clearly restrictive in that it refers to the person Jan Wolkers in his capacity as a writer (and not as a sculptor). However, it seems reasonable to consider this example not as a binominal construction, but as the restrictive counterpart of the appositive construction in (167b).

Examples such as (168) may seem close to the examples in (167), but may be crucially different, since the phrase following the proper noun may simply function as a surname. This is represented orthographically by writing N2 with a capital letter, and may also be reflected in N2 losing its descriptive content.

We also find cases such as Manneken Pis (the name of the famous statue in Brussels) and Vadertje Cats; cf. the appendix in Schermer (2023) for more examples. Such cases cannot be used as vocatives and do not allow N1 to be used as a bare nominal predicate, which Schermer attributes to the fact that the diminutive suffixes (i.e. –ken and –je) make the noun phrase referential.

As a result of the addition of the proper name, the binominal phrase in (163a) becomes uniquely identifying, even out of context. The same effect can be achieved by a (noun phrase containing a) numeral, as in (169a), where it identifies the referent of the full noun phrase. Something similar happens in (169b&c), where the nouns boek and Jan are not used in their normal denoting function, but as meta-linguistic expressions referring to the word themselves.

Sometimes it is not so clear whether we are dealing with true binominal constructions. Consider the examples in (170), from the days when we still had train tickets. Example (170a) differs from the above examples in that it is not a uniquely referring expression. Furthermore, it is possible to express the same meaning by a postnominal PP. This suggests that the binominal construction is simply a shortened version of the noun phrase with a PP-modifier. Something similar could be said for (170b), which can be seen as the shortened version of (170b').

In cases such as (171), the binominal construction as a whole acts as a proper noun, referring to a particular cabinet, committee, method, etc. The second noun is usually the family name of a person closely related to the referent of the noun phrase as a whole. In such cases, the binominal construction is more or less the same as a compound, which is often indicated by the fact that in writing the two nouns are usually joined by a hyphen (especially when the second noun is a proper name).

The examples in (172), where the second noun phrase has the form of a genitive noun phrase, are clearly relics from the older stages of the language. In present-day Dutch, such noun phrases would normally be realized by a postnominal van-phrase instead of the genitive noun phrase; cf. Section 18.2.2.4, sub I.