The core semantics of geen is that of negation, but although geen forms a syntactic constituent with the noun it precedes, the scope of negation is not necessarily limited to the noun phrase; in the vast majority of cases, the negation in geen takes sentential scope. This is particularly clear from the fact, illustrated in (239), that geen can license negative polarity items like ooitever and ook maar Xany X, because these can only be used in the presence of a structurally superior negative element; note that this holds regardless of whether the geen phrase is an argument, as geen autono car in (239a), or an adjunct, as geen momentno moment in (239b). That it is actually the presence of geen that licenses these negative polarity items is clear from the fact that geen does not alternate with een in (239), although this would be possible in the absence of the negative polarity items.

The examples in (240) and (241) also support the conclusion that geen can take sentential scope. First, notice from the contrast in (240a&b) that the sentential negative adverb niet cannot occur in the clause-initial position; example (240b') shows that this is excluded even if niet is part of a topicalized participial verb phrase.

Crucial to our argument is that the unacceptability of (240b') shows that phrases containing sentence negation cannot be topicalized: since topicalization is usually possible in the case of constituent negation, the degraded status of (241b) supports the conclusion that the noun phrase geen brief expresses sentence negation.

This argument is somewhat weakened, however, by the fact that substituting the stronger form of negation geen enkelenot a single for geen results in full acceptability; we will discuss the latter in more detail in Subsection III. Strengthening the negation by adding an empathic accent to the preposed noun phrase may also improve the result, especially in more or less idiomatic expressions such as Geen letter heb ik geschreven!I have not written anything at all. However, in non-idiomatic examples such as (241b), emphatic stress usually leads to a constituent-negation reading; cf. Subsection B.

The fact that the negation takes scope outside the noun phrase may also account for the fact that noun phrases preceded by geen are usually plural when used as the subject of an expletive construction. Semantically, this makes sense, because sentence (242b) is simply the negation of sentence (242a), with a bare plural noun phrase triggering plural agreement: ¬(er lopen kinderen op straat).

Note that the examples in (243) show that geen can also occur in singular subjects, but in this case it is not used as sentence negation. Example (243b) is either used in discourse as an explicit denial of the proposition expressed by (243a) or it receives the stronger “not a single” reading; cf. Section 18.1.1.1 and Subsection III for further discussion.

In contrastive contexts, geen can be used as a constituent negator. If the noun phrase is singular, it usually alternates with niet een, as in (244a), but if the noun phrase is plural, the use of geen is the only option, as in (244).

Topicalization of a geen phrase expressing constituent negation, as in (245a), is at least marginally possible, and seems to lead to a better result than topicalization of the negative adverb niet and its associate noun phrase.

The use of geen in contrastive contexts is excluded when the noun phrase functions as the complement of a PP. In fact, Haeseryn et al. (1997:1657) found that geen-phrases only occur as complements of PPs in idiomatic constructions: Dat ongeluk was met geen pen te beschrijvenThat accident was horrendous. We refer the reader to Section 18.1.5.3 for further discussion.

It is crucial that, outside of contrastive contexts such as (244), replacing geen with niet een is in most cases impossible; in a neutral sentence such as (247), it would definitely be odd to use niet een instead of geen. This again shows that geen cannot systematically be treated as a fusion of niet and the indefinite article een, since this would lead to the incorrect expectation that the infelicity of (247) with niet een would be preserved if niet and een were fused into geen.

An alternative to the fusion approach is to assume that geen is simply lexically marked as a sentence negation, while niet should not be analyzed as a sentence negation when immediately followed by an indefinite noun phrase. Section V13.3.2, sub IC, claims that in such cases niet behaves as a focus particle comparable in several respects to ookalso and alleenonly. The meaning attributed to such particles is that they modify a presupposed proposition in the ways indicated in Table 4 (after Dik 1997a), where the exclamation mark indicates contrastive accent.

	(PA)S	modified set PS	expression type
correcting	X	Y	not X, but Y!
expanding	X	X and Y	also Y!
restricting	X and Y	X	only X!
selecting	X or Y	X	X!

The additional notion of correction of an earlier proposition immediately explains the contrast in acceptability in neutral sentences such as (247), and it seems quite appropriate for describing the subtle meaning difference between the use of geen and niet een in (244a), repeated here in a slightly different form as (248); (248a) simply asserts that of two conceivable eventualities, only one is actualized, while example (248b) corrects an earlier presupposition by saying that the actualized event did not involve the arrival of a letter, but the arrival of a parcel.

The difference in meaning between the two examples in (248) strongly suggests that geen simply expresses sentence negation, and that the constituent-negation reading of (248a) is an epiphenomenon resulting from the focus assignment to the head noun; the only genuine case of constituent negation is given in (248b). The additional claim that the constituent negator niet in (248b) is a focus particle also explains the contrast between the two examples in (246), since (249a) shows that the focus particle alleen must also precede the full PP. Furthermore, example (249b) shows that pied piping of alleen under topicalization is only marginally possible, just like pied piping of niet in (245b).

To conclude this subsection, we note that Haeseryn et al. (1997:1658) also propose a fusion analysis for negative quantifiers/adverbs like niemand, niets, nergens. However, they also note that these forms differ from geen in that the positive form immediately preceded by niet and the negative form are normally in complementary distribution. In neutral sentences like those in (250), it is usually the negative quantifier/adverb that is used. This follows from our earlier proposal that niet is a focus particle that can only be used in the counter-presuppositional context of correction.

In contrastive sentences like those in (251), on the other hand, the form with niet is generally used. This would follow if we assume that, unlike noun phrases of the form geen + N, negative quantifiers/adverbs cannot be used in contrastive contexts. This difference is probably related to the fact that the contrast always involves the nominal head of the phrase geen N, since the quantificational part of the quantifiers/adverbs may be semantically too weak to express contrast.

We conclude that empirical evidence suggests that the fusion analysis of negative quantifiers/adverbs should also be abandoned.

Subsection B suggested that the constituent-negation reading of cases with geen results from the interaction of sentence negation and focus assignment to the head noun. This type of analysis predicts that the scope reduction of the sentence negation expressed by geen need not stop at the border of the noun phrase with geen. Indeed, this is what we find; the scope of the sentence negation expressed by geen can be further reduced to e.g. an attributive modifier within the noun phrase. An ambiguous example of this type is given in (252a), together with its semantically more transparent alternates with niet in the (b)-examples; (252b) has only a reading where niet negates the adjective geringe, while (252b') has only a reading where niet negates the entire noun phrase.

Ambiguity of a similar kind to that in (252a) can be found in noun phrases of the type illustrated in (253a), whose ambiguity comes out in the paraphrases in the (b) examples. Depending on the precise analysis of the internal structure of noun phrases of the type professor Van Riemsdijk (cf. 17.1.3), either (253b) or (253b') instantiates a case in which geen takes scope over a subpart of the noun phrase in which it occurs.

We conclude this subsection with two remarks. First, it should be noted that the ambiguity of (252a) and (253a) exists only on paper, since the intonation will provide clues as to the intended reading. Second, if one were to treat geen in (252a) as the fusion of niet and the indefinite article een, one would have to assume that the order of niet and een is irrelevant; both een niet in (252b) and niet een in (252b') should be able to “fuse” into geen.

Although Subsection C has shown that the scope of the sentence negation expressed by geen can be reduced to elements with which it does not form a constituent, such syntax/semantics mismatches are certainly not possible in just any context. To see this, first observe that the negative adverb niet in (254a) can be construed semantically with the manner adverb goedwell, even though it does not form a constituent with it, which is clear from the fact that it must be stranded under topicalization. In (254b), on the other hand, the negation expressed by the negative quantifier geen cannot be associated with the adverbial phrase; the sentence is at best marginally acceptable on a highly marked count-noun reading of hitte, as in Ik verdraag geen enkele ?(soort) hitte goed I cannot stand any (kind of) heat.

We conclude from this that the scope reduction of the sentence negation expressed by geen is not free, but tied to its noun phrase containing geen in the sense that the negation cannot be semantically associated with other constituents of the clause.

Haeseryn et al. (1997: §29.4) suggest that the negative article geenno is formed by the fusion of the negative adverb niet not and the indefinite article. This subsection has argued on empirical grounds that this analysis cannot be sustained; geen is simply a lexical form (probably not even an article, but a numeral or quantifier) that inherently expresses sentence negation. The fact that geen N sometimes seems to receive a constituent-negation reading similar to that expressed by sequences of the form niet [een N] is not an inherent property of geen, but of the construction in which it is used; in particular, contrastive focus may result in a reduction in the scope of the sentence negation expressed by geen.

The negative quantifier geen sometimes expresses a meaning stronger than simple negation, which we will refer to as the “not a single” reading; cf. Klooster (2001b:57). This reading requires that geen be followed by some contrastively accented element, and can sometimes be strengthened by the addition of certain elements, such as the modifier enkelsingle or the cardinal numeral éénone.

Given the meaning expressed, it is not surprising that the head noun in the “not a single” reading is usually a singular count noun; cf. Postma (1999). However, Hoeksema (2013) correctly points out that there are some cases where geen enkele is followed by a non-count noun, as in (260).

Hoeksema also notes that cases with a plural noun, such as zorgen in (261a), are also possible. He suggests that this may be the result of an ongoing change, but it is not obvious that this is indeed the case because zich zorgen makento worry is a fixed collocation which does not allow the singular form zorg; cf. *zich zorg maken. Since the more or less synonymous construction in (261b) is clearly degraded with the plural noun problemenproblems, we conclude that there is no reason to assume that the noun can be plural in the normal case, although we have to add that occasional cases of this form can be found on the internet.

The use and distribution of “not a single” phrases with geenno, geen enkelenot a single, and geen éénnot one differ in subtle ways. We start with the “not a single” reading of noun phrases with geen. This reading is common for noun phrases with geen in subject position, as in (262a); all subjects with geen in regular object position (i.e. in non-expletive constructions) are of this type. Objects with geen can also receive this interpretation, and for topicalized objects this reading is in fact the only one available; cf. the discussion of (241).

Prosodically, the “not a single” reading of geen phrases is immediately recognizable by the fact that there is main stress on the element immediately following geen. This is often the head noun, but when an attributive adjective is present, it is usually the adjective that receives the greatest prominence.

As in the more usual cases, geen takes scope outside the noun phrase, as can be seen from the fact that it can license negative polarity items such as ooitever; cf. Subsection IA. This is illustrated in (264) for a subject and a topicalized object. We will see, however, that the sentence negation expressed by geen can also receive a reduced-scope reading, in which negation is construed with the accented element following geen.

The “not a single” reading of geen phrases is also common for subjects of comparative constructions.

Within this class of constructions, however, a distinction should be made between comparatives such as in (265), where particular makes or brands of the same type of product are compared, and those such as in (266), where two different types of product are compared. In contrast to the primeless examples, the primed examples in (266) sound distinctly odd.

The assertions that the primed examples in (266) intend to express can be expressed by adding the modifier enkel(e)single, as in (267a&b). In accordance with the generalization that the main stress must be assigned to the element following geen, the main prosodic prominence is assigned to the modifier: geen enkel(e) N.

The modifier enkele can also be used in contexts where reference is made to specific entities (or rather the absence of them). In (267) and (268), the sentence negation expressed by geen clearly receives a reduced-scope reading, since it must be construed with the contrastively focused modifier enkel(e)single; thus, the effect of adding the modifier is that the “not a single” reading is expressed in a more transparent and explicit way.

The reduced-scope reading may also explain why geen enkele N phrases differ from ordinary geen N phrases in that they cannot be used as predicates in complementive constructions, as shown in (269): in (269a), geen has sentential scope and can therefore simply express that Jan does not have the property of being a soldier, but the reduced-scope reading found with geen enkele makes this impossible.

The “not a single” reading can also be emphasized by adding the element één, as in the examples in (270). Here, too, the sentence negation expressed by geen has a clearly reduced scope, since it must be construed with the contrastively focused numeral éénone. The reduced-scope reading is further supported by the fact that één cannot replace enkele in example (269b); cf. *Jan is geen één soldaat.

Observe that the examples in (270) alternate with the constructions in (271) with the focus particle niet in the examples, which is also construed with the contrastively accented numeral één.

The “not a single” interpretation of geen is the one normally found in the numerous idiomatic negative polarity constructions with geen phrases. The idiomatic noun phrases in (272) have the prosody characteristic of the “not a single” cases discussed above: the main accent is assigned to the element following geen. The primed examples show that the idiomatic examples behave like the non-idiomatic ones in allowing topicalization. For a recent corpus study to cases of this kind, see Van den Heede & Lauwers (2024).

The addition of enkel(e)single and éénone is impossible in these idiomatic examples; however, geen can often be intensified by the addition of accented schwa-inflected eneone, as in (273a). In contrast, geen ene cannot easily be used in non-idiomatic examples such as (273b); the uninflected form één or the modifier enkel(e) is highly preferred in this case.

That we are dealing with negative polarity phrases is clear from the fact that the nominal phrase geen (ene) jota/moer can only be replaced by its positive counterpart without geen when the negation is expressed by some other (c-commanding) negative constituent in its minimal clause, as illustrated in (274), which is only acceptable with the negative subject niemand.

The geen-phrases with a “not a single” reading discussed in this subsection are characterized by the fact that they are headed by a singular count noun and contain a contrastive accent. They differ from neutrally accented geen-phrases in that they can occur as the complement of a preposition, as shown by the contrast between the two unprimed and primed examples in (376), taken from Klooster (2001b: §6).

We will not discuss Klooster’s syntactic analysis of the contrast between regular geen-phrases and geen-phrases with a “not a single”, but only note that it is built on two premises: (i) negative noun phrases cannot be directly selected by prepositions (this is ruled out by a filter *P DP[+Neg]); this restriction is lifted when the noun phrase is focused (indicated by the obligatory accent found in the earlier examples). If these premises are correct, they would also account for the acceptability of the examples in (276), in which we find a “not even” reading.

For completeness’ sake, Klooster notes that pronouns such as niemandnobody can be used in PPs because they are always focused in such cases: dat Marie op niemand wil wachten that Marie doesn't want to wait for anyone.

A number of constructions with geen exhibit so-called negative concord, i.e. the multiple occurrence of negative elements with a single negative interpretation as their combined effect; unlike in cases of double negation, there is no cancellation of the negation. These constructions occur only in the spoken language, and some of them may not be part of the standard variety.

The use of geen can invoke evaluative semantics on noun phrases that are otherwise not evaluative in nature, as illustrated in (282a&b). The negation of levenlife by geen leads to an interpretation according to which an emphatically negative evaluation is attributed to life, alternatively expressible by combinations of an adjective and a noun (either compound or phrasal), as in the primed examples.

Geen phrases of this kind only occur in predicative position or as the complement of hebbento have. This is illustrated by the examples in (283): while (283a&b) have an evaluative interpretation, this is not the case in (283c).

Some idiomatic examples can be found in (284). Example (284b) differs from the previous examples in that it involves a positive evaluation: the combination of geen and the substance noun functions as an idiomatic expression meaning “not a small thing, quite something”. Such cases are quite close to litotes, i.e. cases in which the negation is used to emphatically express the opposite of what is expressed by the negated element; cf. Dat is niet niksThis is quite something.

Measure phrases of time and distance, like tien minutenten minutes in (285a) and tien kilometerten kilometers in (285b), can be combined with geen to yield an interpretation that can be paraphrased as “less than X”. The adverbial element nog is typically present alongside geen in such cases, though it seems that it is not strictly necessary in all cases; while in (285a) it would be awkward to omit nog, in (285b) it does not seem completely impossible.

An interpretively somewhat different case of the same type is given in (286) from the Woordenboek der Nederlandsche Taal (item geen). Here the combination of geen and the numeral duizend has an interpretation that can be paraphrased as “not even a thousand”.

In many of the examples in the previous subsections, the core meaning of geen as a negative quantifier seems to be lost. A particularly striking illustration of this fact is provided by examples of the type in (287), where the accent does not fall on geen.

That these are not negative questions is evident from the fact that the speaker who asks a question of the type in (287a) anticipates a positive answer. This is explicitly acknowledged in the answer in (288a) by the addition of the adverb inderdaadindeed. A negative answer is of course possible, but not anticipated by the speaker, which is clear from the fact that the inclusion of the adverb inderdaad in the response in (288b) is pragmatically awkward.

Note, however, that the answer in (288b) with inderdaad is only out of place as a response to (287a) if this question is assigned the prosodic contour typical for questions of this type, with the main accent on courgettes followed by an acutely rising intonation; there is also a truly negative interpretation for (287a), in which geen receives heavy accent, for which (288b) with inderdaad does count as a pragmatically felicitous response.

On the intended, non-negative interpretation of the examples in (287), geen seems dispensable; the examples in (289) can be used in the same contexts as non-negative (287), and are equally acceptable/felicitous. The main difference seems to be that the speaker does not anticipate a positive answer to his question.

We conclude this subsection by pointing out that the negative adverb niet exhibits the same behavior as geen in that it can appear in non-negative questions. In (289), the negative adverb niet can be added to the immediate right of dat with preservation of meaning: Zijn dat niet courgettes? or Is dat niet een leuk idee?