The adjectival amplifiers in (15) are similar to the adverbs in (11) in that they only have an amplifying effect; their original meaning, which is given in the glosses, has more or less disappeared.

Like the adverbs in (11), the amplifiers in (15) cannot themselves be amplified, nor can they occur in negative clauses, as shown in (16).

We can probably include the evaluative adjectives in (17) in the same group as the adjectives in (15): the examples in (18) show that they also cannot be amplified or occur in negative clauses.

Note that (17d) is somewhat special in that the amplifier verrekt can be intensified by the addition of an -e ending, as shown in (19a). Since the AP verrekte moeilijk is used in predicative position, the schwa ending of verrekte obviously cannot be an attributive ending; in fact, the use of the additional schwa has a degrading effect when the AP is used in attributive position, as shown in (19b).

Note that verrekt is used not only as a degree modifier, but also as an evaluative attributive modifier, as in die verrekte opgavethat damned exercise; cf. A1.3.5. It may be that the degree reading of verrekte in (19b) is disfavored by the competition of the fully acceptable stacked reading “that damned, difficult exercise”.

The adjectival amplifiers of the second group have more or less retained the meaning they have in attributive or predicative position. As a result, it is sometimes quite difficult to give a satisfactory translation in English. Some examples are given in (20) and (21).

The use of the adjectival amplifiers in (20) and (21) is highly productive, although it should be noted that they cannot be used to modify an adjective of the same form. This is illustrated in (22).

Note that there are also adjectival modifiers that have fully retained their lexical meaning, but whose main function is not amplification; cf. Section 30.5. Some examples are given in (23).

The main semantic difference between the two sets of amplifiers in (20) and (21) is that the amplification is less strong in the former than in the latter: the amplifiers in (20) express that the state denoted by the modified adjective holds to a high degree, while the amplifiers in (21) express that the state holds to an extremely high or even the highest degree. In other words, the amplifiers in (20) are more or less on a par with the prototypical amplifier zeervery, while the amplifying force of the amplifiers in (21) exceeds the amplifying force of zeer. This can be made clear by the frame in (24a), where the element zelfseven requires the second AP to scale upward with respect to the first; cf. the discussion of (10). Since the amplifiers in (20) cannot be used felicitously in this frame, we can conclude that their amplifying force does not exceed the amplifying force of zeer. On the other hand, the fact that the amplifiers in (21) can be used successfully in this frame shows that their amplifying force is stronger than that of zeer. The use of the number sign again indicates that the string zelfs erg ziek can be used in other contexts.

a.		NPi is zeer A; NP is very	pronouni	is zelfs is even	MODIFIER	A.
b.	$	Jan is zeer ziek. Jan is very ill	Hij he	is zelfs is even	erg very	ziek. ill
c.		Gebouw B is zeer lelijk. building B is very ugly	Het it	is zelfs is even	afgrijselijk atrociously	lelijk. ugly

This difference between the amplifiers in (20) and (21) is also reflected in their gradability. The examples in (25) show that the amplifiers in (20) can themselves be amplified by the adverbs in (11) and undergo comparative/superlative formation.

The examples in (26), on the other hand, show that amplification of the amplifiers in (21) is excluded and that the same holds for comparative and superlative formation.

Apparently, their ultra-high degree reading, triggered by these adverbs, blocks further amplification. Note that the unacceptability of the examples in (26) is not due to some idiosyncratic property of the adjectives; modification and comparative/superlative formation are both possible in the attributive and predicative use of these adjectives, as shown in (27).

Finally, the amplifying force of the amplifiers in (20) can often be increased by reduplication; this is categorically excluded for the amplifiers in (21). This is shown in (28).

Although there seem to be differences between the individual members of the two sets of amplifiers in (20) and (21), they all seem to be possible in negative clauses; cf. also Section 25.3, sub I. This is shown in (29) and (30).

The amplifier ergvery is special because in colloquial speech it can optionally receive an attributive -e ending when modifying an attributive adjective ending in -e; we have seen that the same is possible with the adverb heel in (12a). This is not easily possible with the other adjectival amplifiers, although they do occur in speech. This contrast is illustrated in (31); note that enorme is perfectly acceptable when interpreted as an attributive adjective modifying the noun phrase donkere kamer, i.e. under the interpretation “an enormous/huge dark room”. We refer the reader to Verdenius (1939) and Van der Meulen (2019) for further examples.

In the combination of the adverb heelvery and the adjectival modifier erg, the adverb must precede the adjective, which produces the following inflectional possibilities; the percentage sign in (32b) indicates that speakers seem to differ in their judgments on this example.

Some adjectives are morphologically amplified, as in the case of complex adjectives like beeldschoongorgeous (lit.: statue-beautiful), doodengreally scary (lit.: death-scaring), oliedomextremely stupid (lit.: oil-stupid) and beregoedterrific (lit.: bear-good); the first part of the compound expresses the amplification. As shown in (36), these complex adjectives cannot be modified by additional downtoners or amplifiers.

However, further amplification can often be obtained by reduplication of the first morpheme, as in (37). If the first morpheme of the compound is monosyllabic, the use of the coordinator enand seems to be preferred. If the first morpheme is disyllabic, the reduplicated morphemes can be separated by a comma intonation. We are dealing with tendencies here, as can be seen by the fact that all the forms in (37) can be found on the internet.

The compounds in (36) are generally idiomatic; the first member of the compound cannot be used productively to form intrinsically amplified adjectives. No other compounds can be formed with the amplifying morphemes in (36): *olieschoon, *doodschoon, *bereschoon, *beelddood, #oliedood, *beredood, *beelddom, *dooddom, *beredom, *beeldgoed, *doodgoed, *oliegoed. It is usually assumed that the possible combinations are listed in the lexicon as separate lexical items. It should be noted, however, that around the year 2000 the amplifying affixes dood- and bere- suddenly became quite popular. As a result, some of the starred examples are now easily found on the internet, and may indeed have entered the lexicon as lexicalized forms for a large group of speakers. More recently, the amplifying prefixes super-, ultra-, and mega- have become quite popular, showing that the morphological process of amplification is an easy locus of language variation/change.

The primeless cases in (38) show that amplification can also be expressed by the meer dan A construction, which involves the comparative form of the adjective veelmuch/many. The primed examples show that the comparative form of weiniglittle/few cannot enter into a similar construction.

It is not entirely clear what the internal structure of the predicative phrases in the primeless examples is. Normally it is the comparative that functions as the semantic head of the construction, which is clear from the fact that the dan-phrase can be omitted: cf. Jan is aardiger (dan Peter)Jan is nicer (than Peter). In the primed examples in (38), on the other hand, it is the adjective in the dan-phrase that acts as the semantic head, which is clear from the fact that omitting the dan-phrase yields an uninterpretable result. Therefore, it seems reasonable to assume that the semantic difference should be reflected in the syntactic structure of the predicative phrase; we leave this to future research.

Finally, it should be noted that the adjectives in the primed examples in (39) express more or less the same meaning as the comparative meer dan A construction in the primeless examples.

This shows that some scales can be relevant for more than one adjective, as shown in (40) for intelligent and geniaal in (39a).

The ultra-high degree adjectives in the primed examples in (39) can be amplified, but this easily leads to an ironic or hyperbolic connotation; modification by a downtoner often does not produce a very felicitous result either, and the same goes for comparative/superlative formation. The forms marked with a percentage sign can be found on the internet; the comparative and superlative forms are relatively common in religious contexts, suggesting that they belong to the more formal register.

example	amplification	downtoning	comparative	superlative
geniaal brilliant	%zeer geniaal very brilliant	??vrij geniaal rather brilliant	??genialer more brilliant	??het geniaalst the most brilliant
uitstekend excellent	%zeer uitstekend very excellent	*vrij uitstekend rather excellent	*uitstekender more excellent	*het uitstekendst the most excellent
zalig delicious	%zeer zalig very delicious	*vrij zalig rather delicious	%zaliger more delicious	%het zaligst the most delicious

The equative construction in (42a) compares two properties: it states that the length and width of the table are equal. Note that this example does not imply that the table in question is long or wide; it may be quite short and narrow. This shows that the measure adjectives langlong and breedwide; are neutral in the sense of Section 23.3.2.2, sub ID. This is also supported by the fact that the use of the non-neutral forms kortshort and smallnarrow in (42b) leads to a marked result.

However, if we compare two adjectives of which at least one is not a measure adjective, the implication is that both properties exceed the neutral norm: (43a) implies that Jan is both quite old and quite cunning, and (43b) implies that Jan is both quite intelligent and quite crazy. The constructions in (43) are thus amplifying in nature. See Section 26.1.4 for further discussion of these constructions.

The comparative constructions in (44) both express that Jan is taller than Marie. However, example (44a) need not express that Jan is tall; he can actually be quite short. Similarly, (44b) need not express that Marie is short; she can be quite tall.

If we add the adverb nogeven to the examples in (44), as in (45), the meaning changes radically. Example (45a) expresses that both Jan and Marie are (quite) tall, and (45b) expresses that both Marie and Jan are (quite) short. In other words, the addition of nog has an amplifying effect.

Recall that the comparative form of the ultra-high degree adjective geniaalbrilliant is usually somewhat marked, as indicated in (46a). This markedness disappears completely when the comparative is preceded by nog, as in (46b).

The examples in (45) and (46b) suggest that nog selects a high or ultra-high degree adjective to amplify its degree to an even more exceptional level.

An amplifying effect can also be obtained by accentuating (and lengthening) the adjective, as in the primeless examples of (47), an effect that can even be enhanced by reduplication of the adjective, as in the primed examples. The (a)-examples are complementives, the (b)-examples are attributive adjectives, and the (c)-examples are adverbially used adjectives.

The examples in (47) are exclamative; other exclamative constructions can have a similar amplifying effect. This is illustrated in (48) for exclamative constructions with the exclamative marker wat, which will be discussed in more detail in Subsection V.

The same is true for the exclamative dat constructions in (49). In these constructions, the (phrase containing the) adjective is immediately followed by a clause introduced by the complementizer datthat with the finite verb in clause-final position. The construction as a whole cannot be used as a clausal constituent.

For completeness, note that dat is not a relative pronoun in the attributive (b)-examples; the neuter relative pronoun dat cannot take a non-neuter noun phrase such as een aardige vadera nice father as its antecedent; furthermore, replacing dat with the neuter relative form die leads to unacceptability.