• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
38.2. Monosyndetic, polysyndetic and asyndetic coordination
quickinfo

Coordinators are not lexical heads in the technical sense that they project, i.e. they do not take arguments and cannot be modified. That coordinators are not lexical heads is also reflected in the fact that they are neither unique nor obligatory within coordinate structures: although such structures prototypically contain only a single coordinator before the final coordinand, the coordinator can sometimes also be repeated before all coordinands except the first, or be omitted altogether. The resulting phrases are called monosyndetic, polysyndetic and asyndetic coordinate structures, respectively.

77
a. [Jan, Marie en Peter] komen morgen.
monosyndetic
  Jan Marie and Peter come tomorrow
  'Jan, Marie and Peter will come tomorrow.'
b. [Jan en Marie en Peter] komen morgen.
polysyndetic
  Jan and Marie and Peter come tomorrow
  'Jan and Marie and Peter will come tomorrow.'
c. [Jan, Marie, Peter] ze komen morgen allemaal.
asyndetic
  Jan Marie Peter they come tomorrow all
  'Jan, Marie and Peter, they will all come tomorrow.'

The three types of coordinate structure in (77) do not alternate freely: we will show that the distribution of poly- and asyndetic coordinate structures is more restricted than that of monosyndetic structures. The discussions in subsections I and II will therefore start from monosyndetic coordination and compare it with asyndetic and polysyndetic coordination, respectively. Subsection III concludes with a brief discussion of monosyndetic coordination with a correlative coordinator, as in zowel Jan als Marieboth Jan and Matie because such cases are not always easy to distinguish from polysyndetic coordination with a simple coordinator.

readmore
[+]  I.  Monosyndetic and asyndetic coordination

This subsection discusses some differences between monosyndetic and asyndetic coordinate structures. Monosyndetic coordination is illustrated again in (78) for coordinate structures with more than two coordinands and the coordinator enand; the number of coordinands is in principle infinite (although there are coordinators such as maarbut which allow at most two coordinands). The coordinator is placed before the last coordinand, and the other coordinands except the first are preceded by a short pause, indicated in writing by a comma: since the coordinands must be accented (cf. Section 38.1, sub V), this results in an intonation pattern similar to that found in cases of counting (one, two, three, ...); cf. Bakker (1968:61).

78
a. [Jan, Marie en Peter] komen morgen allemaal.
  Jan Marie and Peter come tomorrow all
  'Jan, Marie and Peter will all come tomorrow.'
b. [Jan, Marie, Peter en Els] komen morgen allemaal.
  Jan Marie Peter and Els come tomorrow all
  'Jan, Marie, Peter and Els will all come tomorrow.'

Asyndetic coordination is illustrated in (79) for the same coordinands found in (78). There is no (phonetically realized) coordinator, and the coordinands are all separated by a short pause, which again results in an intonation pattern similar to that found in cases of counting.

79
a. [Jan, Marie, Peter], die komen morgen allemaal .
  Jan Marie Peter dem come tomorrow all
  'Jan, Marie, Peter they will all come tomorrow.'
b. [Jan, Marie, Peter, Els], die komen morgen allemaal.
  Jan Marie Peter Els dem come tomorrow all
  'Jan, Marie, Peter, Els they will all come tomorrow.'

The main difference between monosyndetic and asyndetic coordinate structures concerns their distribution: Subsection A will show that while the distribution of monosyndetic coordinate structures is not restricted in any obvious way, asyndetic coordinate structures are usually not clause-internal, i.e. they are more or less restricted to main-clause external positions; Subsection B discusses a number of possible counterexamples to this claim. Asyndetic coordinate structures are prototypically interpreted as conjunctions, but Subsection C will show that there are more possibilities in the case of asyndetically coordinated clauses.

[+]  A.  Asyndetic coordinate structures do not occur clause-internally

The most striking difference between monosyndetic and asyndetic coordinate structures concerns their syntactic distribution: while the monosyndetic structures in (78) function as the subject of the clause, the asyndetic structures in (79) are contrastive, °left-dislocated phrases. This is no coincidence, as can be seen from the fact that the primeless examples in (80) are unacceptable; asyndetic coordinate structures simply cannot be used as clausal constituents in colloquial speech. However, it is not the case that asyndetic and monosyndetic constructions are in complementary distribution; the primed examples in (80) show that the asyndetic structures in (79) can easily be replaced by their monosyndetic counterparts. We conclude that monosyndetic coordinate structures have a wider distribution than their asyndetic counterparts.

80
a. * [Jan, Marie, Peter] komen morgen allemaal.
  Jan Marie Peter come tomorrow all
a'. [Jan, Marie en Peter], die komen morgen allemaal .
  Jan Marie and Peter dem come tomorrow all
b. * [Jan, Marie, Peter, Els] komen morgen allemaal.
  Jan Marie Peter Els come tomorrow all
b'. [Jan, Marie, Peter en Els], die komen morgen allemaal.
  Jan Marie Peter and Els dem come tomorrow all

Note in passing that the primeless examples in (80) are acceptable if the coordinate structures as a whole are followed by an intonation break; this is not relevant here, since these examples should then be analyzed as topic-drop constructions with left dislocation, i.e. as constructions with a phonetically empty but syntactically present subject; cf. Section V11.2.2 for a discussion of topic drop. The examples in (81) show that asyndetic coordinate structures cannot occur as a subpart of clausal constituents either. This means that monosyndetic and asyndetic coordinate structures differ in that the latter cannot be embedded in larger syntactic structures (clauses and other phrases).

81
a. De lezing werd [door [Jan, Marie *(en) Peter]] gepresenteerd.
  the talk was by Jan Marie and Peter presented
  'The talk was presented by Jan, Marie and Peter.'
b. [De lezingen [van [Jan, Marie *(en) Peter]]] waren allemaal interessant.
  the talks by Jan Marie and Peter were all interesting
  'The talks by Jan, Marie and Peter were all interesting.'

The above discussion strongly suggests that asyndetic coordinate structures must be main-clause external in the sense of Chapter 37. This correctly predicts that asyndetic coordinate structures can be used in both left-dislocated and right-dislocated positions. That has already been shown for contrastive left-dislocation constructions in (79), and is further illustrated in the (a)-examples in (82) by the hanging-topic constructions with a resumptive subject pronoun in non-initial position of the main clause. The (b)-examples show that asyndetic coordinate structures are also correctly predicted to be able to occur in right-dislocated position: we give these examples as embedded clauses to show that the coordinate structures follow the finite verb in clause-final position, i.e. in a position that cannot be occupied by nominal arguments of the clause. Example (82c) is added to show that asyndetic coordinate structures can also be used as appositions (in parenthetical position); cf. Section N16.1.3.

82
a. [Jan, Marie, Peter], allemaal komen ze morgen.
  Jan Marie Peter all come they tomorrow
  'Jan, Marie and Peter, they will all come tomorrow.'
a'. [Jan, Marie, Peter, Els], allemaal komen ze morgen
  Jan Marie Peter Els all come they tomorrow
b. dat ze morgen allemaal komen: [Jan, Marie, Peter].
  that they tomorrow all come Jan Marie Peter
  'that they will all come tomorrow: Jan, Marie and Peter.'
b'. dat ze morgen allemaal komen: [Jan, Marie, Peter, Els].
  that they tomorrow all come Jan Marie Peter Els
c. Mijn klasgenoten – Jan, Marie, Peter – komen allemaal.
  my classmates Jan Marie Peter come all
  'My classmates —Jan, Marie, Peter— will all come.'

Note that there are (seemingly) asyndetic coordinate structures that can occur in the right periphery of the clause such as Heeft iemand nog vragen, opmerkingen?Does anyone have any questions or comments? We will not digress here into this potential problem for our claim that asyndetic coordinate structures must be in a main-clause external position; we will return to it in subsection C.

We have seen above that monosyndetic coordinate structures have a wider syntactic distribution than asyndetic coordinate structures in that the latter are always main-clause external. It is worth noting that the two types induce slightly different readings when they occur in the same position: asyndetic coordinate structures are open-ended in the sense that they can function as incomplete enumerations, which is also evident from the fact that they are often pronounced with a somewhat wavering intonation. This effect is even clearer in (83a), where the phrase noem maar op!you name it! emphasizes the open-endedness of the enumeration. Monosyndetic coordinate structures, on the other hand, are usually exhaustive, which is why the use of the phrase noem maar op! in (83b) is not felicitous.

83
a. Iedereen was er: Jan, Marie, Peter, ...; noem maar op!
  everyone was there Jan Marie Peter call prt prt.
  'Everyone was present: Jan, Marie, Peter, ... You name it!'
b. $ Iedereen was er: Jan, Marie en Peter, ...; noem maar op!
  everyone was there Jan Marie and Peter call prt prt.

This enumerative use is also common with asyndetically coordinated clauses, as shown in (84), where the last three clauses form a (possibly incomplete) list of reasons for assuming that Jan was ill.

84
Jan was ziek: hij zweette, hij rilde, hij was misselijk.
  Jan was ill he sweated he shivered he was sick
'Jan was ill: he sweated, he shivered and he felt nauseous.'

Finally, note that the generalization that asyndetic coordinate structures do not occur as clausal constituents makes it easy to distinguish such structures from noun phrases with a nominal appositional phrase of the type discussed in Section N16.1.3, such as Marie, onze burgemeester in (85a). The two noun phrases constituting the subject in this example also differ from coordinands in asyndetic (as well as other) coordinate structures in that they do not refer independently: the second noun phrase instead specifies more precisely the entity referred to by the proper name. A foolproof test to distinguish the two cases is to insert the coordinator enand before the second noun phrase. This is possible in asyndetic coordinate structures without any significant syntactic or semantic effect. However, it eliminates the appositional structure and replaces it by a coordinate one, which is reflected both syntactically and semantically: the appositional structure Marie, onze burgemeester in (85a) refers to a single person and triggers singular subject-verb agreement, while the coordinate structure Marie en onze burgemeester in (85b) refers to two different persons and triggers plural agreement.

85
a. Marie, onze burgemeester, is/*zijn erg bekwaam.
appositional structure
  Marie our mayor, is/are very competent
  'Marie, our mayor, is very competent.'
b. Marie en onze burgemeester zijn/*is erg bekwaam.
coordinate structure
  Marie and our mayor are/is very competent
  'Marie and our mayor are very competent.'
[+]  B.  Potential counterexamples

Subsection A has argued that asyndetic coordinate structures have a more restricted distribution than their monosyndetic counterparts. They occur only main-clause externally, i.e. they cannot be used as (parts of) clausal constituents. The claim that asyndetic coordinate structures are main-clause external is not unproblematic. First, such structures can easily be found as clausal constituents in more elevated or literary style (e.g. for metrical or stylistic reasons), which is of course not sufficient to conclude that this option is also part of the core syntax. Second, there are also a number of potential counterexamples that can easily be found in colloquial speech. The following subsections discuss some cases of the second type.

[+]  1.  Numerals

Dik (1997:190) mentions cases of the type in the primeless examples in (86), where two numerals seem to be asyndetically coordinated. Such cases are easy to find in colloquial speech; they differ from the corresponding structure with the coordinator ofor in that a true disjunctive reading is excluded. Examples like this have an approximative reading (“about four or five”), which is especially clear in examples involving higher numbers, as in the primed examples, where the numerals provide a lower and an upper bound (“about thirty to forty”). We can probably set aside these cases because the meaning makes it clear that they do not involve ordinary coordination. This conclusion can be further supported by the fact that the numerals can also be linked by the element à, which is normally not analyzed as a coordinator (but as a preposition): cf. Er wonen daar vier à vijf mensen.

86
a. Er wonen daar [vier, vijf mensen].
  there live there four five people
  'Four or five people are living there.'
a'. Er waren [dertig, veertig mensen] aanwezig.
  there were thirty forty people present
b. Dit nummer duurt [vier, vijf minuten].
  this track lasts four five minutes
  'This track lasts for four or five minutes.'
b'. De sonate duurt [dertig, veertig minuten].
  the sonata lasts thirty forty minutes
[+]  2.  Stacked attributively used adjectives

Haeseryn et al. (1997:1523) mentions the case of stacked attributively used adjectives like those in the primeless examples in (87). It is argued that these adjectives are asyndetically coordinated because they receive a similar (conjunctive) interpretation as those in the predicatively used monosyndetic coordinate structures in the primed examples.

87
a. Peter is een zachtaardig, lief kind.
  Peter is a gentle sweet child
a'. Peter is [zachtaardig en lief].
  Peter is gentle and sweet
b. Marie is een lange, dunne vrouw.
  Marie is a long skinny woman
b'. Marie is [lang en dun].
  Marie is tall and skinny

The claim that we are dealing with asyndetic coordination in the primeless examples is based entirely on semantic grounds, but it is not clear that it is also justifiable on syntactic grounds. For example, the (b)-examples in (88) show that N(P)-ellipsis of the type discussed in Section A27.4 cannot affect a subpart of the attributively used coordinate structure korte en dikke; strikethrough is used to indicate ellipsis.

88
a. Jan wil [een [[lange en dunne] stok]] maar ...
  Jan wants a long and thin stick but
b. ... ik heb alleen [een [korte en dikke] [stok]].
  I have only a short and thick stick
  'Jan is looking for a long, thin stick but I only have a short, thick one.'
b'. * ... ik heb alleen [een [korte en dunne] [stok]].
  I have only a short and thin stick

Example (89) shows that cases with supposedly asyndetically coordinated adjectives behave differently, since the (b)-examples are both fully acceptable. We conclude that the adjectives are not coordinated, but are independent modifiers in a layered structure: [een ... [A1 ... [A2 ... [ ... N ...]]]]; this allows us to omit both N and A2 without affecting a subpart of a coordinate structure.

89
a. Jan wil [een [lange [dunne [stok]]]] maar ...
  Jan wants a long thin stick but
b. ... ik heb alleen [een [korte [dikke [stok]]]].
  I have only a short thick stick
  'Jan is looking for a long, thin stick but I only have a short, thick one.'
b'. ... ik heb alleen [een [korte [dunne [stok]]]].
  I have only a short thin stick
  'Jan is looking for a long, thin stick but I only have a short one.'

Examples such as (89b') are not conclusive for the claim that stacked attributive adjectives must be hierarchically ordered, since we cannot a priori exclude that the object een lange, dunne stok is structurally ambiguous, but these examples at least cast some doubt on the appropriateness of a coordination analysis for (87b); we might as well be dealing with an attributive construction of the type in (90a), which Haeseryn et al. denies an analysis of asyndetic coordination because it would be better paraphrased as in (90b).

90
a. de lage witte huizen
  the low white houses
b. de witte huizen die laag zijn
  the white houses that low are
  'the white houses that are low'

We conclude that, in the absence of conclusive syntactic evidence, we should reject the asyndetic coordination analysis of the primeless examples in (87). Another reason for doing so is that the semantic argument in favor of this analysis is inconclusive anyway, since Section 38.1, sub VII, has shown (among other things) that semantic conjunction can be syntactically expressed by means other than coordination.

[+]  3.  Other cases: complex onomatopoeias and reduplication

Corver (2015a/2015b) mentions another possible counterexample to the claim that asyndetic coordinate structures cannot occur as clausal constituents. Example (91a) shows that onomatopoeias can be used clause-internally. Interjections are often analyzed as clause-external elements because they occur in a position before the main-clause initial position, as shown in (91b). However, it does not seem impossible to place an onomatopoeia in main-clause initial position; example (91c) is marked in comparison to (91b), but seems grammatical.

91
a. De auto reed boem tegen de muur aan.
  the car drove boem against the wall aan
  'The car drove —bang!— into the wall.'
b. Boem, [de auto reed tegen de muur aan].
  boem the car drove against the wall aan
c. (?) [Boem reed de auto tegen de muur aan].
  boem drove the car against the wall aan

Corver shows that complex onomatopoeias should be analyzed as syntactic units. For instance, the examples in (92) show that they can be used as an answer to a wh-question, that they can be used as the complement of “quotative” van (cf. Section V5.1.2.4, sub III), and that they can be coordinated themselves. He also argues that these syntactic units cannot be seen as lexical units (complex words) because the constituent parts do not occur in a fixed order; he concludes that they should be analyzed as coordinate structures.

92
a. Q: Hoe klonk het geluid van de botsing? A: [Pats, boem, knal].
  how sounded the noise of the collision pats boem knal
  'What did the collision sound like?'
b. Jan reed [zo van [pats, boem, knal]] tegen de muur.
  Jan drove so of pats boem knal against the wall
c. Het servies viel [[pats, boem, knal] en [bam, boem, beng]] op de vloer.
  the crockery fell pats boem knal and bam boem beng on the floor

If Corver’s conclusion is correct, the fact illustrated in (93) that complex onomatopoeias such as pats boem knal behave similarly to the simplex ones in (91) would count as a counterexample to the claim that asyndetic coordinate structures do not function as clausal constituents.

93
a. De auto reed [pats, boem, knal] tegen de muur aan.
  the car drove pats boem knal against the wall aan
b. [Pats, boem, knal], [de auto reed tegen de muur aan].
  pats boem knal, the car drove against the wall aan
c. (?) [[Pats, boem, knal] reed de auto tegen de muur aan].
  pats boem knal drove the car against the wall aan

Corver also gives several reasons why the impression given by (91c) and (93c) that the onomatopoeias can function as clausal constituents may be misleading. For instance, the examples in (94) show that onomatopoeias differ from manner adverbials in that they cannot be the antecedent of the proform zoso.

94
a. * Jan reed boem tegen een muur aan en Els reed zo tegen een boom aan.
  Jan drove boem against a wall aan and Els drove so against a tree aan
b. * Jan zakte krak door het ijs en Els zakte zo door de houten vloer.
  Jan fell krak through the ice and Els fell so through the wooden floor

If we can conclude from this that onomatopoeias do not contribute to the logical meaning of the clause, we may also have to conclude that they do not function as clausal constituents. Corver therefore suggests that (91c) and (93c) have an elided pronominal form daar in main-clause initial position as a result of some kind of topic drop and that the onomatopoeias are external to the main clause.

95
a. Boem, [(?)(daar) reed de auto tegen de muur aan].
  boem there drove the car against the wall aan
b. [Pats, boem, knal] [(?)(daar) reed de auto tegen de muur aan].
  pats boem knal there drove the car against the wall aan

Another potential problem for the claim that asyndetic coordinate structures cannot occur clause-internally are the cases of reduplication given in (96), which are used to express adjectival amplification (i.e. very A). The italicized phrases clearly have the syntactic functions given in square brackets.

96
a. De vergadering was saai, saai, saai.
complementive
  the meeting was boring boring boring
b. Je bent een domme, domme, domme jongen!
attributive
  you are a stupid stupid stupid boy
c. Jan liep vlug, vlug, vlug de trap op.
adverbial
  Jan walked fast fast fast the stairs up
  'Jan walked up the stairs very quickly.'

If the cases in (96) could be shown to involve asyndetic coordination, then we would be dealing with uncontroversial counterexamples to our claim. An asyndetic analysis might be in order for (96a), since it can also be realized as a monosyndetic coordinate structure of the complementive in (97a), especially when the modifier nog eens is present; but a similar argument cannot be made for the attributive and adverbial cases.

97
a. De vergadering was [saai, saai, en nog eens saai].
  the meeting was boring boring and again once boring
b. * Je bent een [domme, domme en nog eens domme] jongen!
  you are a stupid stupid and again once stupid boy
c. * Jan liep [vlug, vlug en nog eens vlug] de trap op.
  Jan walked fast fast and once again fast the stairs up

It seems too early to conclude from the examples discussed in this subsection that the otherwise robust generalization that asyndetic coordinate structures cannot occur as (subparts of) clausal constituents is invalid. For the time being, therefore, we assume that asyndetic coordinate structures occur only in main-clause external positions.

[+]  C.  Asyndetically linked clauses

Asyndetic coordinate structures can usually be replaced by monosyndetic ones without any significant effect on the semantic interpretation. In the prototypical case asyndetic coordinate structures are purely conjunctive, as indicated by the fact that they can almost always be replaced by syndetic ones with the coordinator enand. However, if the asyndetically linked coordinands are sentential, they can also be contrastive, and in such cases the coordinator maar can also be used. Some examples, adapted from Haeseryn et al. (1997:1522), are given in (98).

98
a. Soms is ze bangig, (en/maar) soms is ze roekeloos.
  sometimes is she anxious and/but sometimes is she reckless
  'Sometimes she is anxious, (and/but) sometimes she is reckless.'
b. Enerzijds is ze bangig, (en/maar) anderzijds is ze roekeloos.
  on.the.one.hand is she anxious and/but on.the.other.hand is she reckless
  'On the one hand she is anxious (and/but) on the other hand she is reckless.'
c. Hij is bangig, (en/maar) zij is roekeloos.
  he is anxious and/but she is reckless
  'He is anxious, (and/but) she is reckless.'

In the interrogative constructions in (99), also inspired by Haeseryn et al. (1997:1522), the relation between the asyndetically linked clauses is disjunctive nature, as indicated by the fact that the coordinator ofor can be added. Note that the addition of of in (99b) has an interpretive effect similar to that noted above for the addition of en: the asyndetic coordinate structures are open-ended in the sense that they leave more options open, whereas the coordinate structures with of present the addressee with exclusive binary choices (unless the intonation indicates that the sentence is open-ended).

99
a. Zeg het maar: ga je mee? blijf je thuis?
  tell it prt go you prt. stay you home
a'. Zeg het maar: ga je mee of blijf je thuis?
  tell it prt go you prt. or stay you home
  'Well, tell me: are you coming with me or are you staying at home?'

Note that the intonation patterns of the two forms in (99) are also different: in the asyndetic cases, the two interrogative coordinands are pronounced with their own question contours, whereas in the syndetic cases we are dealing with a single question contour: compare Ga je mee? Blijf je thuis? and Ga je mee of blijf je thuis?, where the question mark indicates the rise in tone at the end of the utterance typical of questions. Note that the examples in (100) are probably disjunctions of reduced questions, i.e. [[Kom je vandaag] of/Ø [kom je morgen]]?; cf. Section V5.1.5 on fragment clauses.

100
a. Wanneer kom je hier: vandaag? morgen?
  when come you here today tomorrow
b. Wanneer kom je hier: vandaag of morgen?
  when come you here today or tomorrow
  'When will you be here: today (or) tomorrow?'

Ellipsis may also provide a solution to the problem mentioned in Subsection A that examples such as (101a) are fully acceptable. We cannot analyze the two noun phrases as the coordinands of an asyndetic coordinate structure, as in (101b), because this is incompatible with our earlier conclusion that asyndetic coordinate structures cannot be used as clausal constituents (here: direct object). However, we can explain the acceptability of (101a) by assuming that we are dealing with a sentential coordinate structure, with a fragment (i.e. reduced) clause as the second coordinand. This analysis seems to be supported by the fact that (101a) can be comfortably pronounced with a single question contour when ofor is present, but not when it is absent: compare Heeft iemand nog vragen of opmerkingen? with Heeft iemand nog vragen? Opmerkingen? where the question marks again indicate the typical rise in tone at the end of the utterance.

101
a. Heeft iemand nog vragen (of) opmerkingen?
  has someone still questions or remarks
  'Are there any questions or remarks?'
b. Heeft iemand nog [vragen of/*Ø opmerkingen]?
b'. [[Heeft iemand nog vragen] of/Ø [heeft iemand nog opmerkingen]]?
[+]  II.  Monosyndetic and polysyndetic coordinate structures

Polysyndetic coordination is illustrated in (102b&c) for coordinate structures with three and four coordinands, respectively, and the coordinator enand; the number of coordinands is in principle infinite. Each of the coordinands except the first is preceded by a coordinator; the omission of any of these coordinators precludes polysyndetic analysis of these coordinate structures. The polysyndetic coordinate structures in (102b&c) differ from their monosyndetic counterparts in (78) in that the coordinators are emphatically accented (which is indicated by small caps); for this reason it seems justified to consider example (102a) with the indicated intonation pattern also as a case of polysyndetic coordination, despite the fact that there is no more than one coordinator present.

102
a. [Jan en Marie] komen morgen.
  Jan and Marie come tomorrow
b. [Jan en Marie en Peter] komen morgen allemaal.
  Jan and Marie and Peter come tomorrow all
  'Jan and Marie and Peter will all come tomorrow.'
c. [Jan en Marie en Peter en Els] komen morgen allemaal.
  Jan and Marie and Peter and Els come tomorrow all
  'Jan and Marie and Peter and Els will all come tomorrow.'

Polysyndetic conjunctive coordination differs from monosyndetic conjunctive coordination in that its semantic use seems to be more restricted: while (103a) allows either a distributive or a cumulative interpretation in the sense that the persons referred to by the subject may each have lifted the rock individually or may have lifted the rock together, example (103b) seems to prefer a distributive interpretation, although Dik (1968: §11) and De Vries (2005) claim that a cumulative reading is also possible. The semantic contrast between the two examples suggested here seems to be confirmed by the fact that the addition of the modifier samentogether, which triggers a cumulative reading, leads to a marked result in the case of a polysyndetic coordinate structure; cf. Section 38.1, sub IVD.

103
a. [Jan, Marie en Els] hebben de rots opgetild.
distributive/cumulative
  Jan Marie and Els have the rock prt.-lifted
  'Jan, Marie and Els have lifted the rock.'
a'. [Jan, Marie en Els] hebben samen de rots opgetild.
cumulative
  Jan Marie and Els have together the rock prt.-lifted
  'Jan, Marie and Els have lifted the rock together.'
b. [Jan en Marie en Els] hebben de rots opgetild.
distributive preferred
  Jan and Marie and Els have the rock prt.-lifted
  'Jan and Marie and Els have lifted the rock.'
b'. ?? [Jan en Marie en Els] hebben samen de rots opgetild.
cumulative
  Jan and Marie and Els have together the rock prt.-lifted
  Compare: 'Jan and Marie and Els have lifted the rock together.'

The examples in (104) show that polysyndetic coordination is also possible with the disjunctive coordinator ofor. The polysyndetic coordinate structures in (104b&c) again differ from their monosyndetic counterparts in that the coordinators are accented; (104a) with the given intonation pattern can therefore also be considered a case of polysyndetic coordination, despite the fact that there is only one coordinator.

104
a. [Jan of Marie] komt morgen.
  Jan or Marie comes tomorrow
b. [Jan of Marie of Peter] komt morgen.
  Jan or Marie or Peter comes tomorrow
c. [Jan of Marie of Peter of Els] komt morgen.
  Jan or Marie or Peter or Els comes tomorrow

The generic examples in (105) suggest that disjunctive polysyndetic coordination is also semantically more restricted than its monosyndetic counterpart. The former is more strictly exclusive than the latter: example (105a) can be used to express that biographies, detective stories and thrillers constitute Jan’s reading material, while (105b) expresses that Jan’s reading material is limited to only one of these genres (although the speaker does not know which one), or perhaps that Jan goes through successive phases in which he reads only one of these genres.

105
a. Jan leest biografieën, detectives of thrillers.
inclusive disjunction
  Jan reads biographies detectives or thrillers
  'Jan reads bibliographies, detective stories or thrillers.'
b. Jan leest biografieën of detectives of thrillers.
exclusive disjunction
  Jan reads biographies or detectives or thrillers
  'Jan reads biographies, detective stories or thrillers.'
[+]  III.  A note on correlative coordinators

Subsection II has shown that all non-initial coordinands in polysyndetic coordinate structures like (106a&a') are preceded by a coordinator. The fact that the first coordinand is not preceded by a coordinator distinguishes polysyndetic coordinate structures from coordinate structures like (106b&b') with a correlative coordinator, since the first part of the correlative must precede the initial coordinand. The two types of coordinate structures in (106) are difficult to distinguish for other reasons: they both require an accent on the coordinators (as indicated by small caps), and they both receive a distributive interpretation in the sense that the persons mentioned are involved in different eventualities of “being on vacation”.

106
a. [Jan en Marie] zijn op vakantie.
polysyndetic
  Jan and Marie are on holiday
a'. [Jan en Marie en Peter] zijn op vakantie.
  Jan and Marie and Peter are on holiday
b. [En Jan en Marie] zijn op vakantie.
correlative coordinator
  and Jan and Marie are on holiday
b'. [En Jan en Marie en Peter] zijn op vakantie.
  and Jan and Marie and Peter are on holiday

The fact that the two types differ only in the presence of an “initial” coordinator may raise the question whether it is justified to distinguish the two structures. The answer should be affirmative, since we also have correlative coordinators like zowel ... als ...both ... and ... and (formal) hetzij ... of ...either ... or ..., where the two parts of the correlative coordinator are not homophonous.

107
a. [Zowel Jan als Marie] is op vakantie.
  both Jan and Marie is on holiday
  'Both Jan and Marie are on holiday.'
b. De bijeenkomst zal [hetzij morgen of donderdag] plaatsvinden.
  the meeting will either tomorrow or Thursday] place-take
  'The meeting will take place either tomorrow or on Thursday.'

Coordinate structures with correlatives are always “polysyndetic” when there are more than two coordinands; the examples in (108) show that replacing the coordinator (part) before the second or third coordinand with an intonation break yields a degraded result (although the percentage sign is used in (108b) to indicate that taaladvies.net/taal/advies/vraag/1223 claims that als can be replaced by a comma); the same is true for structures in which both coordinators are replaced by an intonation break; cf. *[Zowel Jan, Marie, Peter] is op vakantie.

108
a. [En Jan *(en) Marie en Peter] zijn op vakantie.
  and Jan and Marie and Peter are on holiday
a'. [En Jan en Marie *(en) Peter] zijn op vakantie.
  and Jan and Marie and Peter are on holiday
b. [Zowel Jan %(als) Marie als Peter] is op vakantie.
  both Jan and Marie and Peter is on holiday
  'And Jan and Marie and Peter are on holiday.'
b'. [Zowel Jan als Marie *(als) Peter] is op vakantie.
  both Jan and Marie and Peter is on holiday

The (b)-examples also show that it is the second part of the correlative coordinator that is repeated before the non-initial coordinand(s); substituting zowel for the first occurrence of als leads to unacceptability regardless of the form of the element preceding the initial coordinand, as shown by the unacceptability of the examples in (109).

109
a. * [Zowel Jan zowel Marie als Peter] is op vakantie.
  both Jan both Marie and Peter is on holiday
b. * [Als Jan zowel Marie als Peter] is op vakantie.
  and Jan both Marie and Peter is on holiday

This suggests that the initial part of the correlative has a special status. This will be confirmed by the more detailed discussion of correlative coordinators in Section 38.4.2, where it will be argued that the initial part is not a coordinator-like element, but a focus particle.

References:
    report errorprintcite