PhiN 24/2003: 27

Konstanze Jungbluth (Erfurt/Tübingen)

How do speakers of Creoles broaden their vocabulary?
NN-sequences in Romance-based Atlantic Creole languages

How do speakers of Creoles broaden their vocabulary?
NN-sequences in Romance-based Atlantic Creole languages

This paper focuses on word formation, more precisely, on noun-noun-sequences in Portuguese-based Crioulo de Cabo Verde. Furthermore the data of Papiamentu and Palenquero, both Spanish-based Creoles, and of French-based Ayisyen spoken on Haiti are taken into consideration. The word formation patterns are compared to the former superstrata languages, too. By adding to the universal and African substrate dependent perspectives the Romance approach one-sided assessments will be straighten out and our understanding of word formation in Creole Languages will be deepened.1

1 Introduction

The goal of this research paper is threefold: First I'd like to question the widely held view that Creole languages are simple and primitive. By presenting not only words or sentences but in fact a short text sample in Cape Verdean Creole, I hope that the reader can best develop his/her own intuitions about the functioning of a Creole. Second, I have singled out from the data striking aspects of noun-noun-sequences, on which I will show the orientation of the Creole to European, that is Romance, patterns2. One of the most important tasks for the speaker of a Creole is to continually develop and normalize diverse possibilities of expression with a relatively limited primary vocabulary, in order to be able to do justice to the various necessities of daily and festive communication. To compare to the data on the Portuguese-based Cape Verdean Creole3, I will take into consideration the Spanish-based Creoles Palenquero and Papiamentu4, as well as the French-based Creole Ayisyen5. Third, I consider this to be a contribution to research of Creole languages from the Romance perspective. Most Creole language research is dominated by studies that are done in the generative paradigm. They are mostly interested in sentence structures. Data from Creoles are used in the hope of learning more about human-created Universal Grammar, as proposed by these linguists.6 Other researchers, familiar with one or another African language, choose another important perspective of Creoles. They will recognize African patterns in the Creole data, and place the influence of the substrate language in the foreground.7 But supporting this work is a comparison of Old and New World Romance-speaking areas. I will consider the Creole data against the background of the recognizably Romance pattern. In my opinion, we can deepen our understanding and straighten out one-sided assessments by considering them from three viewpoints, that is, the addition of universal and African substrate dependent perspectives to the Romance perspective.

This work is composed of 4 parts: the Introduction; Composition Patterns in the Romance Superstrata Languages (French, Spanish, and Portuguese); Noun-noun Sequences in Creole Languages contains two parts, Directly and Prepositionally Connected Noun-Noun Sequences in Crioulo do Cabo Verde and the comparison of the pattern found therein to Palenquero, Papiamentu and Ayisyen; and finally, The Contribution of Romance Linguistics to Creole Studies.

PhiN 24/2003: 28

The four different Romance-based Creoles under consideration are shown on the following map.

PhiN 24/2003: 29

The origins of these Atlantic Creoles can be understood as a product of the overseas trade triangle between Europe, Africa and America. The names of the cities marked on the map owe to this historical context. Creoles have established themselves primarily on islands. The communicative necessities of a multilingual society accompany the development of these comparatively young languages. The plantation economy, particularly the cultivation of sugar cane, required a constant source of manpower, which above all meant the import of African slaves to America. Also comparable to the situation on the islands are the slave settlements on the South American continent, which lay in inaccessible terrain. Cimarrones, slaves who escaped from the plantations and mines, constitute the beginning of these speech communities. Palenquero, represented by 1 on the map, is spoken by about 3000 people in San Basilio, Colombia. Papiamentu, represented by 2 and likewise Spanish-based, is used by about 200,000 speakers on the so-called ABC islands, the former Dutch Antilles, and in fact has official status.8 Spoken on Haiti, the French-based Ayisyen9 also has co-official status. With about 7 million speakers, the Haitians represent the largest Creole-speaking society in the Caribbean region. In the center of this work is Crioulo do Cabo Verde, which is Portuguese-based and is spoken by about 500,000 inhabitants of the seven islands archipelago off the coast of Senegal. Even more speakers than on Cabo Verde itself are scattered throughout the world, such that it is possible to talk of an about one million member linguistic culture. Larger Creole-speaking cultures live in North America, especially in Boston, in Holland, and on the west coast of Africa. The vast majority of studies available are of English-based Creoles. The French-based are also comparatively well studied. Far behind are the other, Spanish or Portuguese-based Creoles. Crioulo do Cabe Verde is the oldest known Creole still in use, and is in this respect of particular interest for linguistics. The historical development of the language reaches back almost 500 years. The islands were uninhabited until 1460, the year of their 'discovery' by the Portuguese.

Now that the linguistic situation has been detailed, I'd like to turn again to the research question: How do the Creoles broaden their vocabulary? This task is fundamental in the context of the history of a language which, by necessity, had to take vocabulary from former or concurrently-used trade languages. It may surprise the reader that the treatment of Creoles did not begin from this point.

Talvez a composição seja o único processo morfológico que se encontra em todos os crioulos e até mesmo nos pidgins. Eu creio que seria de grande valor teórico um estudo comparativo do processo morfológico composicional nos diversos crioulos do mundo. É bem provável que a composição seja o processo morfológico mais primitivo, no sentido de primordial, primevo ou primeiro na evolução linguística.10 (Couto 1994: 83, emphasis K.J.)

PhiN 24/2003: 30

My research paper demonstrates a framework in which this research suggestion can correspond to compositional procedures with regard to Crioulo do Cabo Verde, whereby it is possible to compare this study with the first findings on Romance-based Creoles, which were recently elaborated upon in Paris and Curaçao. Nomen-Nomen-sequences (NN-sequences) have until recently been little researched, even though in this context they acquire a particular importance. The literature is therefore dedicated to generalities, because the noun is more general compared to the verb, and, to cite Aristotle:

Zuerst müssen wir feststellen, was Nomen und was Verbum, dann,was [...] Aussage und Rede ist. (Aristoteles 1920: 1)

This task is not at all trivial in the context of Creole research, as I will show. Specific to one language I support myself with the recently published Dicionário do Crioulo da Ilha de Santiago, which relied on the research efforts of the DFG-Project11 on Crioulo under the direction of Jürgen Lang in Erlangen. I will also take under consideration a few of his transcribed folktales and anecdotes, which allowed me to analyze the data in its context.

2 Patterns of composition in the Romance superstrata languages

In order to bring the background of Old World Romance-speaking area into focus, I have chosen two examples from each of the relevant Romance base-languages to demonstrate the pattern of Romance binominal composite word formation. Compared to the familiar and frequent composites of German, for example Schlüsselloch (keyhole) or Wortbildung (word formation), this method is much less common among Romance speakers. Nonetheless it can occur in two forms: one, as a direct noun-noun connection, in which the order of the determiner and the determined is the reverse of that in German. Spanish bocallave, literally mouth + key for 'keyhole', shows this postdetermining order. The other form that such noun-noun connections in Romance languages show is the connection with the preposition de. In diachronic perspective, the Latin genitive case marking is replaced by the prepositional additive de. Looking at the words' formation, some of these combinations solidified into composites that represent an object in spoken reality. For example, libro de bolsillo, 'pocket book', is a bound composite that generally does not permit insertions into its internal structure.

Table 1: Patterns of Noun Compositions in the Romance Superstrata Languages

Composite French Spanish Portuguese
N+Prep+N livre d'images – photo album edición de bolsillo12 – pocket book porquinho-da-índia – (piglet+PREPDET+ India) guinea pig; cavy
  épreuve de journal – (test+de+newspaper) galley (proof) formación de palabra – word formation formação de palavra – word formation
NN pause café – coffee break bocallave13 – (mouth + key) keyhole beira-mar14 – (bank + sea) beach
  wagon-lit – sleeper [car] hombre-rana – (man + frog) frogman pontapé – (point + foot) kick

PhiN 24/2003: 31

Writing the compound with and without hyphen, together or separated is only secondary for the assessment of the composite character of the word formation. It mostly indirectly shows that the order of these binominal sequences in Romance languages between word formation and syntagma show fluid transitions.

Since data from Creoles has a particular character, I have chosen the term "NN-Sequences" and not "noun composites" in the subtitle of this work, the latter would be premature at this stage of research. The data must thus be interpreted in multiple ways, and its segmentation must first be clarified. Which sound combinations are allowable as words? Second, the question must be answered as to whether words involved are really nouns or not, because a morphologic marking on the word stems themselves is not common. Usually one and the same sequence of sounds can represent a noun, adjective or verb. The transcribed data correspondingly do not show any markings as to which part of speech the words should be ascribed to. Their classification is created by the context, be it a conversational context or the situation itself. In addition the decision must be made in a further step as to whether the syntagmatic order should be considered a compound or as a syntagma alone. Even in this point it is necessary to weigh different aspects of form and content in order to prepare a relevant classification.

3 NN-sequences in Creoles

The following text excerpt, in which may be found various noun determinations but in which many NN-sequences may be observed, should provide an integrated impression of the Portuguese-based Atlantic Creole we are studying. The excerpt is drawn from a corpus of 44 transcribed anecdotes, which André dos Reis Santos gathered from an elderly Cape Verdean by the name of Nastásia. She tells a story typical to this cultural context, about a young man who emigrated to America, and then, once of marriageable age, returned. Before he can ask the hand of his beloved in the house of his future in-laws, he must first undergo a metamorphosis, a kind of complete re-acculturation.

Illustration 1: NN-sequence pánu bicu

1 Kel otu dia, N la na kása Serban15
  DP: that A: other N: day PP1SG.UNBET: I V: go ADV: there PREP N: house PROPN
  The other day I went off to Serban's store.
2 N kunpra sirola nóbu,          
  PP1SG: I V: buy N: underwear A: new          
  I bought new underwear,
3 N kunpra águ xéru di téra,      
  PP1SG: I V: buy N: water N: fragrance PREP N: land      
  a locally made perfume,
4 N kunpra kamisa tabeladu16          
  PP1SG: I V: buy N: shirt A: pleated          
  a pleated shirt,
5 N kunpra kálsa di pánu bicu17      
  PP1SG: I V: buy N: trousers PREP N: fabric N: animal      
  trousers of a coarse fabric.

PhiN 24/2003: 32

6 Kúsas di Mérka N po di un bánda,  
  N: things PREP PROPN: America PP1SG: I V: lay PREP DETINDEF N: side  
  I set aside the things from America.
7 N bisti so kusas di téra.      
  PP1SG: I V: wear/put on only N: things PREP N: land      
  I only wore local [Cape Verdean] things,
8 Piu18 kusas di téra.          
  only N: things PREP N: land          
  only local things.
9 N aránka pa Len Ferera, dentu bóka tardi19    
  PP1SG: I V: go away PREP PROPN PREP N: mouth N: afternoon    
  For the afternoon I went to Len Ferera. (NL 17/35)

The illustration shows that nouns can be more closely determined via various methods in Creoles as well. The nouns specified by postposed adjectives do not differentiate themselves from Portuguese expressions in terms of sequence (lines 2 and 4). To understand the data in the following, of particular use are the terms di tera, literally 'of the land', meaning 'local'; and the last term boka tardi, by which the interpretation of tardi as a noun is only determined by context, such that this word can also be used as an adjective in the meaning of late and as verb in the meaning of 'to linger', 'to allow oneself time', and also in the sense of 'to come late'.

3.1 Directly and Prepositionally Bound NN-sequences

I would like to focus in particular on the apparently unbound sequences of two nouns, which I call direct NN-sequences.

In order to broaden the base of data, I present another example which demonstrates this commonly occurring sequence. The present tense of these connections in every Creole language text stands out from the norm in texts of Old World Romance languages. Even three nouns are often set beside each other. In the following illustration the word bóka 'mouth', which we know from the previous example, appears again, this time in the middle of the group.

Illustration 2: NNN-sequence nxada bóka gudja

10 Ten nxada bóka20 gudja pa koba koba21 simentera.  
  PRES: there is/are N: axe N: mouth N: needle PREP V: dig N: ditch N: seed  
  There is a pickaxe for digging a furrow. (DCIS: 'nxada')

PhiN 24/2003: 33

The phrase di tera, which appeared in Illustration 1 with the sense of 'local, native, from Cabo Verde' is altered in the following illustration to -l tera, because the preceding noun ends with a vowel and the succeeding noun begins with a consonant. Moreover, the additive sucri-l tera, with which the native (light brown) sugar is meant, depicts a lexical unit, which must be recognized to contain a lemmatic quality that forces the closing together of the nouns. The example originates from the collection of folktales Na bóka noti, literally 'In the Mouth of Night'.

Illustration 3: NN-l N-sequence sucri-l téra

11 Un bes, Lobu ku Xibinhu teneba kada un  
  one time, Lobu with/and Xibinhu have-PAST each one  
  Once upon a time, both Lobu and Xibinhu had
12 d'es un fórma sukri-l téra22 (NB 332/1)      
  of them a form sugar-of land        
  a piece of local sugar.

This close compounding of two nouns also occurs with parts of the body, which in this case take on an intrinsically specific meaning. Their entry in the Crioulo da Ilha de Santiago dictionary illustrates a lexicalized use that no longer varies freely:

Illustration 4: N-l N-sequence kántu-l bóka, N d-N-sequence kántu d'odju

13 kántu-l bóka kántu d'odju kántu-l perna dédu-l pé23
  corner-of mouth corner of-eye corner-of leg finger-of food
  corner of the mouth corner of the eye groin toe

This illustration will be useful later in connection with observations of the Spanish-based Creole Palenquero.

A comparison of the following two words shows how the creative compounding of two two-syllable words in both cases leads to a new lexical unit, which can also be considered a single compositional multisyllabic noun. However, its lemmatic quality must at least be accounted for. From águ, 'water' and odju, 'eye' are formed:

Illustration 5: N d-N Sequences

14 águ d'odju24 odju d'águ
  water of-eye eye of-water
  tear(s) well

PhiN 24/2003: 34

Frequent use stabilizes the convention and assures its entry into the lexicon. In addition, it is not rare to find lexemes built in similar fashion at different times, syntactic aggregations, and possibly loanwords from Portuguese that are either conventional in other speech styles or marked as "high speech". 25

The Iberoromance illustrations above strengthen my decision to conceive the direct NN-sequences as maximally integrated, because they may be taken from a different perspective as new, multisyllabic lexical units in the vocabulary and therefore stand virtually ready for a new process of compounding. I would turn attention back to Illustration 2, in which the noun bóka gudja shows a closer relationship than that with nxada and both following nouns.

The data presented from Crioulo do Cabo Verde can in my opinion be summarily and sensibly placed between the poles of Aggregation and Integration. The conception developed from sentence links seems plausibly fruitful for word compounding as well.

Am einen Extrem, dem Pol der Aggregation, stehen zwei Sätze unverbunden nebeneinander. Am anderen Extrem bleibt ein einziger, völlig integrierter Satz übrig. Dieser einzige Satz ist aber gleichzeitig wieder Ausgangspunkt für das Hinzufügen eines weiteren Satzes – und dabei kann erneut die Reihe der Techniken von links nach rechts durchlaufen werden. Seiler versinnbildlicht die Nähe der beiden extremen Lösungen zueinander mit dem topologischen Sonderfall eines Möbius-Bandes. (Raible 1992: 29)26

Table 2: Noun compound patterns in Crioulo do Cabo Verde

Compound Crioulo do Cabo Verde
NN agu xéru – eau de toilette, perfume
pánu bicu – fabric with animal pattern
N+Prep+N kántu-l bóka – corner of the mouth
águ d'odju – tear(s)

PhiN 24/2003: 35

In preparation for the next section, I would like to stress that the parallelism between form and degree of integration in Crioulo do Cabo Verde can at least be observable to a certain degree, even when isolated examples must be classified at the edges of their respective zones on the Möbius Strip above.

3.2 Crioulo do Cabo Verde in Comparison with Palenquero, Papiamentu and Ayisyen

A cursory look at both individual studies, with regard to the noun complexes in Iberoromance-based Creoles which we are interested in, shows that the Cape Verdean Crioulo's observable variation in direct NN-sequences and in NN-sequences bound by di can also be established in Palenquero, but not in Papiamentu, which in support of the superstrata model regularly binds compounds using di.27

Table 3: Patterns of Noun Compounding in Palenquero, Papiamentu and Ayisyen

  Palenquero Papiamentu Ayisyen28
NN ngláno aló – grain of rice
kabésa ngómbe – cow's head
kabei boto29 – lift (lit: back of the boat) po-bouch – lips (skin + mouth)
izin sikr– sugar factory
mal-tèt30 – headache
N+PREP+N técho i posá – roof31
lélo i pié – toe (lit. finger of the foot)
palu di garganta – 'neck bones'
doló i kabes – headache32
(Dijkhoff 1999: 6)
(Krouwenberg/Murray 1994: 33)

In summary, the four Creoles presented here can be characterized in the following way:

Table 4: Comparison of Noun-Noun-sequences in the Creoles in consideration

  NN N + Prep + N
Papiamentu   +
Crioulo do Cabo Verde + +
Palenquero + +
Ayisyen (Haiti) +  

This observation is surprising given that the French-based Ayisyen also knows no binding at the sentence level, and in this way is different from French. Papiamentu has conventionalized compounding with di to such an extent that comparison with Spanish texts is no longer possible. In no way are both of the Creoles to which both patterns are available based on the same Romance language. Palenquero is based on Spanish, Crioulo do Cabo Verde on Portuguese.

The use of the formation patterns varies specifically to each individual language, such that the combinations observed in Crioulo do Cabo Verde between form and degree of integration cannot be seamlessly transferred into the word formation of the other Creoles in consideration here.

PhiN 24/2003: 36

While the absence of the pattern of prepositional compounds is yet be established for Ayisyen, this area of concern embraces nearly the whole. According to Dijkhoff, the small remainder simply represents the binominally transparent composites loaned from Spanish. For its part, Palenquero shows more variation of the two compounding methods than Crioulo do Cabo Verde. What all Creoles have in common is that a shift toward the Integrative pole is generally accompanied by a joining of the nouns. I interpret the elision of vowels in Cape Verdean and the elision of consonants in Papiamentu and Palenquero in this sense.

4 The Contribution of Romance Linguistics to Creole Studies

The Romanist perspective contributes a deepening of understanding of these methods in research into Creole languages. The methods of binding connectorless NN-sequences, meager from a universal perspective and which in some African languages is preferably used for inherent combinations, converges with both known morphological types from the Romance languages: the availability of the pattern from various sources contributes in my opinion to a high frequency of these bi- and multi-nominal compounds in Creoles33. Research of Creoles must seek a connection between these different perspectives, because

Single-cause theories [...] ignore the important possibility that there may be a conspiracy between the different forces. Most notably, one can expect combinations of

1 superstratum and universal tendencies
2 substratum and universal tendencies
3 substratum and superstratum
4 all three factors.

Such combinations occur at all levels of grammar (Mühlhäusler 1998: 132, emphasis K.J.)

The influence of the substratum with regard to composites is validated by Lefebrve34 (1998) and Moñino35 (to appear).

The patterns of these methods of word formation are, however, just as demonstrable in the superstrata languages. Many of the NN-composites in Romance languages also show a genitive function36. The convergence of the patterns may have led to the high productivity of the NN-sequences in the Creole language data. The complexity of the processes do not differentiate themselves from the established processes of the Old World in this area of word formation. Composites originate, like in all languages, according to the need to communicate needs. On the basis of a relatively uncomplicated primary vocabulary, the Creole speaker coins new words from two or more nouns with predetermined binds, words that over time are normalized. In exactly the same way, the variations within a single language, which do not copy the preferred methods of the respective former colonial languages but rather develop various predilections in the Creoles, show that the patterns of word formation are productive in individually characteristic ways. The comparison of different Romance-based Creoles of the Atlantic region demonstrates the autochthonous development of these languages.

PhiN 24/2003: 37

List of Abbreviations

A Adjektive NVok Noun that ends with a vowel
ADV Adverb PAST Past tense
AUX Auxiliary Verb PL Plural
DADV Adverb of place POSS. Possessive pronoun
DEF Definite PP1. SG. Personal pronoun, 1. Person Singular
DET Article PP2. Pl. Personal pronoun, 2. Person Plural
DP Demonstrative Pronoun PREP Preposition
GLIEDSIG Discourse particle PRES Presentative
IMPF Imperfect PROPN Proper name
INDEF Indefinite TMA Tense-Aspect-Mode-Marker
KONN Connector (un)bet. (un)stressed
MOD Modal verb V Verb
N Noun VMOD Modal verb
NEG Negator VokN Noun that begins with a vowel
NKon Noun that ends with a consonant Vperiph Auxiliary verb in the Verbal periphrase


Aristoteles (1920): Perihermenias oder Lehre vom Satz. Ed. Eugen Rolfes. Leipzig: Meiner.

Azevedo do Campo, José Luís (2000): Portugiesisch-basierte Kreolsprachen in Westafrika und Ostasien – ein Überblick, Universität Rostock (Lehr- und Arbeitshefte der Uni. Rostock)

Bartens, Angela (1996): Der kreolische Raum. Geschichte und Gegenwart, Helsinki: Suomalainen Tiedeakatemia.

Bartens-Adawonu, Angela (1999): "A génese dos crioulos cabo-verdianos por difusão componencial e a importância dos estudos dialectológicos", in: Zimmermann, Klaus (ed.): Lenguas criollas de base lexical española y portuguesa, 67–88.

Bechara, Evanildo (1999): Moderna Gramática Portuguesa. Rio de Janeiro (Lucerna).

Bentolila, Alain (1978): "Créole d'Haïti. Nature et Fonction Naturelle", in: Etudes Crèoles, 1: 65–75.

PhiN 24/2003: 38

Benveniste, Émile (1967/1974): "Fondaments syntaxiques de la composition noune", in: Problèmes de linguistique générale, Vol. 2, Paris (Gallimard), 145–162.

Bickerton, Derek (1981): Roots of Language, Ann Arbor (Karoma).

Bustos Gisbert, Eugenio de (1986): La composicion nominal en Español. Salamanca: Univ. de Salamanca.

Couto, Hildo Honório do (1994): O crioulo portugu&ecric;s da Guiné-Bissau, Hamburg: Buske.

Coseriu, Eugenio (1966): "Tomo y me voy. Ein Problem vergleichender europäischer Syntax", in: Vox Romanica 25, 13–55.

Coseriu, Eugenio (1977): "Inhaltliche Wortbildungslehre", in: Brekle, Herbert E./Kastovsky, Dieter (eds.): Perspektiven der Wortbildungsforschung, Bonn (Bouvier): 48–61.

Creolist Archives:

DCIS siehe Lang, Jürgen (2002)

DeGraff, Michel (in press): "Word formation in Creole Genesis", in: Kenstowicz, M./Hale, K.(eds.): A Life in Language, Cambridge: MIT-Press.

Dijkhoff, Marta B. (1999): "Complex Nouns and Composite Nouns in Papiamentu", in: Maurer, Philippe/Stolz, Thomas, Varia Creolica, Bonn: Brockmeyer, 1–10.

Fleischmann, Ulrich (1986): Das Französisch-Kreolische in der Karibik. Tübingen (Narr).

Gather, Andreas (2001): Romanische Verb-Nomen-Komposita. Tübingen (Narr).

Hagège, Claude (2001): "Creoles and the notion of simplicity in human languages", in: Linguistic Typology 5–2/3, 167–175.

Hundertmark-Santos Martins, Maria Teresa (1982): Portugiesische Grammatik, Tübingen (Niemeyer).

Kouwenberg, Silvia/Murray, Eric (1994): Papiamentu. München: LINCOM.

Lang, Jürgen (1991): "Die Kategorie Numerus im kapverdischen Kreol", in: Neue Romania 10, 1–19.

Lang, Jürgen (1994): "Estruturas eventualmente africanas no crioulo de Cabo Verde (Santiago)", in: Papia 3,2, 171–176.

Lang, Jürgen (2002): Dicionário do crioulo da ilha de Santiago (Cabo Verde).Tübingen: Narr.

PhiN 24/2003: 39

Lefebrve, Claire (1998): Creole Genesis and the acquisition of grammar. Cambridge (CUP).

Lehmann, Christian/Moravcsik, Edith (1999): "Noun", in: Lehmann, Christian/ Skopeteas, Stavros/ Kesselheim, Wolfgang (eds.): HSK Morphologie. Berlin: de Gruyter, 732–757.

Lopes da Silva, Balthasar (1957, reimpr. 1984): O Dialecto Crioulo de Cabo Verde. Lisboa: Casa da Moeda.

McWhorter, John H. (2001): "The world's simplest grammars are Creole Grammars", in: Linguistic Typology 5–2/3, 125–166.

Moñino, Yves (a paraître): Les constructions génitives en palenquero: une sémantaxe euro-africaine, 1–24.

Mühlhäusler, Peter (1986): Pidgin & Creole Linguistics. Oxford: Blackwell.

Munteanu, Dan (1996): El papiamento, lengua criolla hispánica. Madrid: Gredos.

Perl, Matthias (1999): "Problemas actuales de la estandarización del papiamentu", in: Zimmermann, Klaus (ed.): Lenguas criollas de base lexical española y portuguesa, 251–260.

Plank, Frans (2001): "Creoles – a structural type?", in: Linguistic Typology 5–2/3, V.

Raible, Wolfgang (1992): Junktion. Eine Dimension der Sprache und ihre Realisierungsformen zwischen Aggregation und Integration. Heidelberg: Winter.

Rainer, Franz (1993): Spanische Wortbildungslehre. Tübingen (Niemeyer).

Schuchardt, Hugo (1888): "Beiträge zur Kenntnis des kreolischen Romanisch I. Allgemeines über das Negerportugiesische", in: ZrP 12, 242–54. Zur Rezeption im englischsprachigen Raum: Gilbert G. Glenn, Pidgin and Creole Languages. Selected essays by Hugo Schuchardt, Cambridge (CUP) 1980. (!)

Schwegler, Armin (1996): "Chi ma kongo": Lengua y rito ancestrales en El Palenque de San Basilio. Frankfurt a.M. (Vervuert).

Slaby, Rudolf/Grossmann, Rudolf (1975): Wörterbuch der spanischen und deutschen Sprache. 2 Bände, Wiesbaden (Brandstetter).

Thiele, Petra (1991): Kabuverdianu: Elementaria seiner TMA-Morphosyntax im lusokreolischen Vergleich. Bochum (Brockmeyer).

Valdman, Albert (1988): Ann pale kreyòl,. Bloomington (Creole Institute).

Veiga, Manuel (1995): O Crioulo do Cabo Verde: Introdução a Gramática. Praia (Instituto do Livro, Instituto Nacional da Cultura).

PhiN 24/2003: 40


BN Silva, Tomé Varela da (Ed.): Na bôka noti (BN): vol. I, Praia (Institutu Kapverdianu di Libro): 1987, 27–37.
DCIS Lang, Jürgen (2002): Dicionário do crioulo da ilha de Santiago (Cabo Verde): Tübingen: Narr.
MOR Mornas (MOR)
NL Nastási Lópi (NL): anedotas em cassette, transcrição feita por André do Reis Santos, 4, 7, 12, 15, 17.
PRE La Prensa (Newspaper in Curaçao): 9.7.1994 (PRE)


1 I would like to thank Jürgen Lang (Erlangen), Ulrich Fleischmann (Berlin), Hildo Couto (Brasília), Angela Bartens (Helsinki) und Yves Moñino (Paris) for valuable hints and various support. I owe thanks to Aria Adli, Dorothee Kaiser and Paul Gévaudan for their valuable suggestions. The translation was done by Kevin Marston.

2 cp.: Romance Languages in general: Coseriu 1977, Coseriu 1966, Benveniste 1967/1974, Gather 2001; Portuguese: Hundertmark-Santos Martins 1982, Bechara 1999; Spanish: Bustos Gisbert 1986, Rainer 1993.

3 Lang 1991, Lang 1994, Thiele 1991, Veiga 1995, Lopes da Silva (1957, reimpr. 1984); cp. o Crioulo da Guiné-Bissau: Couto 1994.

4 Kouwenberg/Murray 1994, Munteanu 1996, Dijkhoff 1999, Perl 1999; for Palenquero cp. Schwegler 1996, Moñino (to appear).

5 Valdman 1988, Bentolila 1978, DeGraff (in press).

6 Bickerton 1981, McWhorter 2001. cp. Hagège 2001, Plank 2001.

7 Moñino (a paraître).

8 Bartens 1996, Bartens-Adawonu 1999, Azevedo do Campo 2000, Schuchardt 1888.

9 Fleischmann 1986.

10 It may be that composition is the only morphologic process that is to be found in all of the Creoles and in our pidgins as well. I believe that it would be of great theoretic value to comparatively study the compositional morphologic process of the various Creoles in the world. It is very probable that composition is the most primitive morphologic process, in the sense of primordial, primeval or first, in linguistic evolution. Translation: K.J./Kevin Marston.

11 Deutsche Forschungsgemeinschaft: German Research Foundation.

12 Rainer describes these composites as syntagmatic composites: "Im Einklang mit der Tradition sehe ich den Unterschied (zwischen Kompositum und freier Phrase K.J.) in der begriffsbildenden Funktion. [...] Genaugenommen enthalten N+Präp+N-Komposita als Bezeichnungen für Begriffe immer zusätzliche semantisch-enzyklopädische Information, die man erst durch eine Kenntnis des Referenten erwerben kann." (Rainer 1993: 291).

PhiN 24/2003: 41

13 "Beim Typ bocacalle hat das Zweitglied die Funktion einer genitivischen de(+Artikel)+Substantiv-Phrase mit possessiver Bedeutung: 'ein N1 des/eines N2': Diachron betrachtet ist zumindest in einigen Fällen ja auch tatsächlich das de ausgefallen. Pluralisiert wird im allgemeinen das Gesamtwort bocacalles, telerañas. Die meisten usuellen Bildungen haben nur einen Akzent. [...] (Es) dominiert bei den usuellen Bildungen die Zusammenschreibung," (Rainer 1993: 259–260, emphasis K.J.) cp. footnote 36.

14 "É muito natural em português a omissão da preposição de , como acontece com arco-íris (por arco da íris)." (Bechara 1999: 356).

15 "e'ka fla m'el ba kasa saibu " and she didn't tell him that she had gone to a medicine man (lit. house [of a] wise one).

16 cp. span.: tableado – adj.; falda tableada – pleated skirt (Slaby/Grossmann)

17 cp. span. (Rio Plata!): bichará – Poncho of coarse fabric (Slaby/Grossmann); crioulo pánu bitxu – fabric with a kind of relief pattern (animal pattern?) (cp. illust. in Lang DCIS 2002: 539)

18 cp. port.: pio – peeps, nem pio! Not a sound! (Irmen)

19 There is also a Crioulo verb: tardi – to linger, to come too late; and a Crioulo adjective: tardi – late; in the DCIS nouns are differentiated by using a diacritic. Thus tárdi is represented with a diacritic, in order to mark it as a noun, but cp. footnote 13. "Nouns generally seem to form a distinct class by syntactic criteria, although not necessarily by morphological or phonological criteria." (Lehmann/Moravcsik 2000: 733)

20 crioulo: bóka fáka – knifeblade (Lang 2002)

21 in the Original, kóba: Lang differentiates the nouns from homophonous verbs with the accent diacritic. Whether or not this differentiation actually has its basis in phonetic realization is disputed.

22 cp. span.: azúcar piedra – (sugar + rock) 'coarse grained sugar'.

23 Couto's orthography includes a hyphen: Kriôl da Guiné-Bissau: dedu-di-pe (Couto 1994: 84).

24 Lang has águ di odju. I made the change myself, because the phonetic elision of the first vowel makes águ d'odju the much more probable transcription.

25 cp. fonti (portug. fonte) well.

26 "At one extreme, the Aggregation pole, two unconnected sentences stand next to each other. At the other extreme is a single, completely integrated sentence. This single sentence is, however, at the same time another starting point for the addition of a further sentence – and thereby the left-to-right progression of methods can be run through again. Seiler represents the closeness of both extremes to each other with the special topological case of a Möbius Strip." Translation Kevin Marston

27 At the very least this characterization is valid in regard to new constructions: "Apart from simple, base nouns and derivative nouns, there are also some compound nouns in Papiamentu. But the strategy for forming the latter nouns does not seem to be very productive. There are, consequently, not many of them. New words are usually formed by combining simple words with the connective di." (Dijkhoff 1987: 1, footnote 2).

28 This pattern differs from the usual French pattern, to compare:
krey.: Li pitoboulèt moripase boulet vijann.(Bentolila 1978: 71)
fr.:Il préfère les boulettes de morue aux boulettes de viande.
 He prefers fishcakes over eatballs.

PhiN 24/2003: 42

29 Kabes di boto in Dijkhoff. The orthographic representation kabei boto is in accordance with Kouwenberg/Murray's account, 1994: 33, which is possibly attributable to a semantic difference between the two representations: The literal counterpart is in accordance with the three-word sequence, with the meaning 'elevator' having the reduced form.

30 But also:
depi pitit la gen on ti tèt fè mal, yo mennen n' kay manbo.
After the child has a slight headache, they bring him to the healerwoman.(Bentolila 1978: 69)

31 A possessive determination follows: técho i posá mi.

32 * doló di mi kabes is ungrammatical in Papiamentu.

33 "La surprenante similitude du génetif palenquero avec les langues Niger-Congo sans classes comme le gbaya a été soulignée, notamment dans les modalités d'usage des deux formes. [...] les premiers locuteurs du palenquero ont sélectionné di- pour 'copier' la structure génitive ibéro-romane avec de ou di, mais en l'utilisant pour marquer seulement les relations génitives contingentes (e.g. relations non nécessaires)," (Moñino to appear: 20)

34 Lefebvre (1998: 47): "It is hypothesised that the creators of the Creole use the principles of word concatenation of their own grammars in developing the Creole. [...] This explains, why [...] Creole languages follow the patterns of their substratum languages."

35 Moñino (to appear: 7): "Malgré des differénces notables dans ses modalités, le palenquero et le créole (e.g. de Guadeloupe K.J.) expriment une opposition d'autonomie/dépendance dans les relations génitives. Ils connaissent aussi une construction directe par simple juxtaposition, qui peut difficilement être attribuée à des tendances internes du français ou de l'espagnol. […] Sémantiquement, la relation d'equivalence s'oppose à la relation génitive et il est très difficile d'imaginer une extension logique de l'une à l autre dans nos créoles à partir des langues romanes. Passons maintenant aux constructions de la relation génitive dans les langues de l'Afrique sub-saharienne." (emphasis K.J.)

36 NN-composites in Spanish, according to Rainer (1993: 259–263):
Left-headed N+N-composites:

  • actor-bailarín: 'an x, that is simultaneously N1 and N2'
  • hombre-anuncio: 'an N1, that functions as an N2'
  • hombre-rana 'an N1, that looks like an N2'
  • referendum-farsa 'an N1, that is an N2'
  • coche bomba 'an N1 with an N2'
  • silla jineta 'an N1 for an N2'
  • papel aluminio 'an N1, whose material quality is characterized by N2'
  • fútbol-sala 'locative type'

Right-headed N+N-composites:

  • madrepatria.

Note that only the example 'papel aluminio' represents a genitive relation. cp. footnote 5.