to the noun than the true adjectives. For instance, in Rendille we find the order given in (22), (22) sticks-my four long white where the opposite relative adjacency is found. To explain the difference between the two orderings, we must start from the twofold nature of common nouns. This notion derives from the logical analysis of nouns by Bach (1968), who argued that a noun is a variable combined with a predicate; for example, a doll means ‘an x that is a doll’. In Langacker’s (1991: 54) terms, a nominal displays grounding and type specification (among other things). Most languages (such as English) apparently attract first the modifiers that fit in best with the type specification, that is, the predicational content of the noun, whereas other languages (such as Rendille) highlight the grounding aspect by attracting first the deictic modifiers. At the same time, what the two language types have in common is the relative order of the nominal modifiers with respect to each other. Adjacency of Similar Elements A third principle involves the adjacency of similar elements, whereby similar ele- ments are placed together. Thus, in languages with the basic orders SOV and VSO, the nominal elements subject and object are placed together. As such, for instance, in the Dutch subordinate clause in (23), the nouns (proper names) are positioned next to each other. (23) dat John Sheila haat that John Sheila hates ‘that John hates Sheila’ In Dutch, there is also a tendency for verbal elements to cluster in the so-called verbal end-group, even ifthis disruptsthe adjacency ofheadanddependent, as in (24): (24) dat John Sheila uit Londen terug zal brengen that John Sheila from London back will bring ‘that John will bring back Sheila from London’ Although terugbrengen ‘bring back’ is one verb, the nonverbal prefix terug can be separated from its verbal head brengen, as in (24), so that the verbal elements zal and brengen cluster together. However, the order zal terugbrengen is also possible. Again, we see two competing motivations, in this case head proximity as opposed to adjacency of similar elements. Relator in the Middle (RIM) A special iconic principle of adjacency is the one whereby a so-called relator is put in the middle of its two relata. One example of this principle is the typical or- dering in coordination whereby a coordinator is placed between the elements it conjoins, as in John and Mary. There are several such constructions. 410 willy van langendonck a. Relator Constructions. A relator can be defined as a free or bound mor- pheme that has basically two syntagmatic slots, or relata, in its semantic- syntactic structure, such that the relator defines a specific semantic- syntactic relation between the two relata. The first relatum is more gen- eral in nature, has a freer position in the sentence, and can sometimes be dropped; the second relatum is usually obligatory and has a fixed posi- tion because it has a tighter bond with the relator. As a consequence, it is normally not omitted. 17 In the following typology, the nature of the relators and their relata will be dealt with (see also Van Langendonck, Swiggers, and Van de Velde, forthcoming). We can distinguish two major subclasses of relators: coordinative and subordinative relators, with the latter further subdividing into predicative and nonpredicative relators. Coordinative relators include coordinating conjunctions, such as and in (25). Predicative relators are verbs and other predicates, such as loves in (26); nonpredicative relators include adposi- tions (i.e., pre- or postpositions), subordinating conjunctions, and certain particles. 18 An example of an adpositional (prepositional) relator is in, as found in (27). Usually, only this third type of relator is explicitly rec- ognized as such in the literature. However, it seems advisable to broaden the definition so as to include the three categories mentioned. (25) John and Mary (26) John loves Mary. (27) The girl (is) in the garden. Coordinative relators are defined by the fact that their two conjuncts (re- lata) mostly seem to show a certain symmetry in that they are inter- changeable, at least in principle. Still, even in productive coordination the first conjunct is more prominent pragmatically and often refers to the el- ement occurring first in time or space. In freezes, where the order of the elements is fixed, the unmarked conjunct acts as the first relatum. Aside from such freezes, the second conjunct displays a tighter bond with the relator, being syntactically more essential: (28) a. John and Mary b. the king and the queen c. here and there d. She got pregnant and married. By contrast, with subordinative relators, the two relata are hardly inter- changeable because of the clearly asymmetric relationship between the two. Again, the first relatum is usually more prominent. According to the function of the first relatum, we can distinguish two types of subordina- tive relators: predicative and nonpredicative. With predicative relators, which are mostly verbs, the first relatum functions as the subject of the iconicity 411 verb. With nonpredicative relators, the first relatum is either the relator’s head or the subject of the predicate of a clause. In both types, the second relatum has again a tighter bond with its relator, functions as a kind of complement, and can hardly be omitted. Let us now consider the two types of subordinative relators more in particular. Nonpredicative relators have a complement as their second relatum: in (29a), for instance, the prepo- sition in heads its object the garden;in(29d), the comparative particle than heads its object Alice. The first relatum is either the relator’s head, as in (29a)–(29c): girl, excitement, did, or the subject of the clause’s predicate, as in the comparative sentence (29d): Kevin. (29) a. the girl in the garden b. the excitement before his departure c. He did it before he left. d. Kevin is taller than Alice. Predicative relators have the subject of the predicate as their first relatum and a complement as their second relatum. In case a verb has several complements, we have to do with more than one ‘‘second’’ relatum (as in 30a). The prototype of a predicative relator is a transitive verb. As is well known, in a sentence such as John killed Bill, the subject John refers to the agent of the action of killing, of which the patient, expressed by the object Bill, is the victim. We take predicative relators to refer here to all sorts of verbal categories (including auxiliaries and modal verbs), as well as predicative adjectives, as in She is worth it. As second relata of main verbs, we consider not only direct objects (as in 26 above) but, for instance, also indirect objects (as in 30a) and prepositional objects (as in 30b): (30) a. It cost me that. 19 b. She looked after him. b. Iconic Ordering in the Relator Construction. As indicated by the principle formulated above, relator constructions often show iconic ordering in that relators take middle positions. Dik (1983: 274) states that ‘‘the preferred position of a Relator is in between its two relata.’’ This syntactic order reflects the fact that the relator establishes a specific semantic connection between the relata. In the iconic ordering, the most prominent relatum takes the first position, whereas the second relatum follows the relator. Notice that it is only when the relator is a full word that it can exert any influence on order. A good example of the principle of ‘‘Relator In the Middle’’ (RIM) is the basic order of subject and object. As is well known, the order SVO (42%) is far more frequent than the order OVS (1%) across languages (Tomlin 1986: 22). Note that SVO competes mainly with SOV 412 willy van langendonck and VSO order, where we find the principle of ‘‘adjacency of similar elements.’’ c. RIM Languages. It appears to be possible to identify ‘‘RIM languages’’ in which relators are typically put in between their relata. Thus, in the North- ern European area we have discovered a belt of languages that we can call RIM languages. These languages show the canonical order in relator con- structions, but the (basic) order modifier-before-head in other construc- tions. As we could see above, English is such a language and forms part of this belt—as do Scandinavian, Finnish, and Russian, among others (see Van Langendonck, Swiggers, and Van de Velde, n.d.). For relator con- structions, I can refer to examples (25) through (30) above. Construc- tions without a relator exhibit [modifier < head] as their unmarked order. 20 This is illustrated in (31). (31) demonstrative < noun: that town predeterminer < demonstrative: all those numeral < noun: three plants adjective < noun: nice girl, red cap proper name modifier < noun/participle: a London shop, the Everard Brothers, Italy based adverb < adjective: extremely intelligent adverb < adverb: very well, not quite compounds: broomstick, furniture shop To conclude these paragraphs on word order, we can say that word-order iconicity appears to be constituted by one general principle of closeness (or, alter- natively, distance), under which come various subprinciples. Three kinds of close- ness have been reviewed: closeness of events in narrative sequence, closeness to the prototypical speaker or to the speakers as a physical entity, and finally, closeness in content. Under the latter, I subsume simple adjacency, relative adjacency, adja- cency of similar elements, and the medial positioning of relators. 3. General Conclusion In this survey of iconicity, I have emphasized diagrammatic iconicity in language and related it to markedness and to the prototypical speaker. It has also become clear that iconicity fits in well with the cognitive and experiential tenets of Cog- nitive Linguistics. Both isomorphism and motivation figure as important phe- nomena throughout such basic cognitivist works as Langacker (1987, 1990, 1991). iconicity 413 NOTES For this overview of iconicity, I have especially benefited from the pioneering work done by John Haiman in numerous publications (e.g., 1980, 1983, 1985). 1. For comments on Peirce in relation to the linguistic sign, see, among others, Pharies (1985). 2. For overviews and reflections on iconicity, see, among others (besides Haiman’s work), Bouissac, Herzfeld, and Posner (1986), Van Langendonck and de Pater (1993), Motivation et Iconicite ´ (1993), and Simone (1994). 3. Certain authors observe that there are problems with these terms and concepts. Greenberg (1995: 57–58) finds the term ‘‘isomorphism’’ unfortunate. Others point out that the notion of ‘‘motivation’’ goes beyond that of iconicity (Motivation et Iconicite ´ 1993). 4. According to Kleiber (1993: 120), the device of schematicity may turn out to be too powerful since there is obviously no limit to the level of abstractness that can be applied to the semantic definitions of the grammatical units in order to preserve isomorphism. 5. This generalized isomorphism paradoxically leads to the introduction of the notion symbolic: ‘‘Lexicon, morphology and syntax form a continuum of symbolic units’’ (Lan- gacker 1990: 1). 6. In certain cases, the generalization of isomorphism generates real conflicts with motivational iconicity. Thus, Kleiber (1993: 121) contends that Langacker’s (1987: 216) analysis of adjectives as relational predicates (like verbs) goes against the iconicity of motivation because the landmark of adjectives is in fact never expressed (see also note 20). I signaled a similar conflict between isomorphism and motivation with regard to proper names (Van Langendonck 2004, 2007). 7. There are several criteria for defining markedness, which may even contradict the ‘‘more form, more meaning’’ criterion, but we cannot go into this here. See, however, Greenberg (1966a) and, for a more recent account, Croft (2003). 8. Compare the notion of ‘‘embodiment’’ (see Rohrer, this volume, chapter 2). 9. Related to Mayerthaler’s (1980) concept of the prototypical speaker is Langacker’s (1985) notion of ‘‘subjectivity’’ as opposed to ‘‘objectivity.’’ Both Mayerthaler’s and Lan- gacker’s ideas lead to an explanation of the so-called animacy or empathy hierarchy, a controversial topic in linguistic typology. 10. Langacker (1991: 447) speaks of a distancing effect of the conjunction that even in pairs like She knows that he likes her versus She knows he likes her. 11. In the same vein, Ruwet (1984) deals with so-called equi-NP deletion: in French we have to say je veux partir ‘I want to leave’ instead of *je veux que je parte ‘I want that I leave’. However, if the volition and the action of the agent are independent of each other, two propositions are necessary and equi-NP deletion cannot apply: je pre ´ fe ´ rerais que moi je puisse faire cela ‘I would prefer I could do that’ (see also Langacker 1991: 448). 12. The iconic difference between kill and cause to die was ignored by Generative Semantics, which derived kill from cause to die by a prelexical transformation. Fodor (1970) already criticized this derivation by pointing out that cause to die, but not kill, may imply a difference in time, for example in John caused Bill to die on Sunday by stabbing him on Saturday; see especially Wierzbicka (1975). 13. Of course, languages may differ as to the rigidity of this cline. For instance, while English may have a car accident killed him, the Dutch equivalent is not acceptable: *een auto-ongeval doodde hem. This makes Dutch more iconic than English in this respect. 414 willy van langendonck Apparently, the subject in English, which grammaticalizes topics rather than agents, allows for more patterns than it does in Dutch, where the subject is primarily characterized by agentivity. 14. Russian tam i sjam ‘there and here’ is an exception; here phonetic iconicity appears to have overridden the semantic principle in that stops tend to precede fricatives in such constructions: t before sj (see also Ross 1980). 15. Of course, politeness or political correctness may change this order, as in ladies and gentlemen. 16. In its most general form, the principle of adjacency was already formulated by Otto Behaghel (1932: 4): ‘‘Das oberste Gesetz ist dieses, dass das geistig eng Zusammengeho ¨ rige auch eng zusammengestellt wird’’ [The primary law is that what belongs closely together semantically is also closely placed together]. Rijkhoff (1992: 214) speaks of a principle of domain integrity. 17. For example, in a sentence like John is in London, where in is the relator, John the first relatum, and London the second, London has a fixed position immediately after in, whereas John does not immediately precede the preposition. 18. The label ‘‘particles’’ also includes ad hoc morphemes that are hard to accom- modate in an ordinary word class, e.g., than. 19. In this instance, the verbal relator has two ‘‘second relata,’’ the direct and the indirect object. 20. That the pattern [adjective þ noun] figures among the nonrelator constructions appears to contradict Langacker’s claim that adjectives are ‘‘relational’’; see also note 6. REFERENCES Anttila, Raimo. 1972. Introduction to comparative and historical linguistics. New York: Macmillan. Bach, Emmon. 1968. Nouns and noun phrases. In Emmon Bach and Robert Thomas Harms, eds., Universals in linguistic theory 90–122. New York: Holt. Bauer, Laurie. 1996. No phonetic iconicity in evaluative morphology: Results of tests on a genetically diverse group of 50 languages. Studia Linguistica 50: 189–206. Behaghel, Otto. 1932. Deutsche Syntax: eine geschichtliche Darstellung, Band 4. Wortstellung, Periodenbau. Heidelberg: Winter. Benveniste, Emile. 1946. Relations de personne dans le verbe. Bulletin de la Socie ´ te ´ Lin- guistique de Paris 43: 1–12. Birdsong, David. 1995. Iconicity, markedness, and processing constraints in frozen locu- tions. In Marge E. Landsberg, ed., Syntactic iconicity and linguistic freezes: The human dimension 31–45. Berlin: Mouton de Gruyter. Bolinger, Dwight. 1977. Meaning and form. London: Longman. Bouissac, Paul, Michael Herzfeld, and Roland Posner, eds. 1986. Iconicity: Essays on the nature of culture: Festschrift for Thomas A. Sebeok on his 65th birthday.Tu ¨ bingen: Stauffenburg. Brinton, Laurel J. 1987. Diagrammatic iconicity in English syntax. Semiotic Inquiry 7: 55–72. Chomsky, Noam. 1972. Language and mind. Enlarged ed. New York: Harcourt Brace Jovanovich. iconicity 415 Chomsky, Noam. 1976. On the nature of language. In Stevan Harnad, Horst Dieter Steklis and Jane B. Lancaster, eds., The origins and evolution of language and speech 46–57. New York: New York Academy of Sciences. Chomsky, Noam. 1980. Rules and representations. New York: Columbia University Press. Chomsky, Noam. 1981. On the representation of form and function. Linguistic Review 1: 3–40. Clark, Herbert H. 1973. Space, time, semantics and the child. In Timothy E. Moore, ed., Cognitive development and the acquisition of language 27–63. New York: Academic Press. Comrie, Bernard. 1981. Language universals and linguistic typology. Oxford: Basil Blackwell. (2nd ed., 1989) Croft, William. 2003. Typology and universals. 2nd ed. Cambridge: Cambridge University Press. Dik, Simon C. 1983. Two constraints on relators and what they can do for us. In Simon C. Dik, ed., Advances in functional grammar 267–98. Dordrecht, Netherlands: Foris Publications. Dobrizhoffer, M. 1902. Auskunft u ¨ ber die abiponische Sprache . Leipzig, Germany: J. Platzmann. Fischer, Olga. 1997. Iconicity in language and literature: Language innovation and language change. Neuphilologische Mitteilungen 98: 63–87. Fodor, Jerry A. 1970. Three reasons for not deriving ‘kill’ from ‘cause to die’. Linguistic Inquiry 1: 429–38. Geeraerts, Dirk. 1997. Diachronic prototype semantics: A contribution to historical lexicology. Oxford: Clarendon Press. Genette, Ge ´ rard. 1976. Mimologiques: Voyage en Cratylie. Paris: Seuil. Givo ´ n, Talmy. 1994. Isomorphism in grammatical code: Cognitive and biological con- sideration. In Raffaele Simone, ed., Iconicity in language 47–76. Amsterdam: John Benjamins. Goossens, Jan. 1969. Strukturelle Sprachgeographie. Heidelberg: Winter. Greenberg, Joseph H. 1966a. Language universals. In Thomas A. Sebeok, ed., Current Trends in Linguistics, vol. 3, Theoretical foundations 61–112. The Hague: Mouton. Greenberg, Joseph H. 1966b. Some universals of grammar with particular reference to the order of meaningful elements. In Joseph H. Greenberg, ed., Universals of Language, 73–113. Cambridge, MA: MIT Press. Greenberg, Joseph H. 1995. On language internal iconicity. In Marge E. Landsberg, ed., Syntactic iconicity and linguistic freezes: The human dimension 57–63. Berlin: Mouton de Gruyter. Hage ` ge, Claude. 1982. La structure des langues. Paris: Presses universitaires de France. Haiman, John. 1974. Concessives, conditionals, and verbs of volition. Foundations of Language 11: 341–59. Haiman, John. 1980. The iconicity of grammar: Isomorphism and motivation. Language 56: 515–40. Haiman, John. 1983. Iconic and economic motivation. Language 59: 781–819. Haiman, John. 1985. Natural syntax: Iconicity and erosion. Cambridge: Cambridge Uni- versity Press. Hamilton, Helen W., and James Deese. 1971. Does linguistic marking have a psychological correlate? Journal of Verbal Learning and Verbal Behavior 10 : 707–14. 416 willy van langendonck Heine, Bernd. 1980. Determination in some East African languages. In Gunter Brettsch- neider and Christian Lehmann, eds., Wege zur Universalienforschung: Sprachwis- senschaftliche Beitr € age zum 60. Geburtstag von Hansjakob Seiler 180–86.Tu ¨ bingen: Gunter Narr. Hetzron, Robert. 1978. On the relative order of adjectives. In Hansjakob Seiler, ed., Language universals 165–84.Tu ¨ bingen: Gunter Narr. Hudson, Richard. 1984. Word grammar. Oxford: Basil Blackwell. Jakobson, Roman. 1965. Quest for the essence of language. Diogenes 51: 21–37. Jakobson, Roman. 1966. Implications of language universals for linguistics. In Thomas A. Sebeok, ed., Current Trends in Linguistics, vol. 3, Theoretical foundations 263–78. The Hague: Mouton. Kakehi, Hisao, Ikuhiro Tamori, and Lawrence Schourup. 1996. Dictionary of iconic ex- pressions in Japanese. In cooperation with Leslie J. Emerson. 2 vols. Berlin: Mouton de Gruyter. Katz, Jerrold J., and Paul M. Postal. 1964. An integrated theory of linguistic descriptions. Cambridge, MA: MIT Press. Kleiber, Georges. 1993. Iconicite ´ d’isomorphisme et grammaire cognitive. Faits de langue 1: 105–21. (Special issue on ‘Motivation et Iconite ´ ’) Lakoff, George. 1982. Categories and cognitive models. LAUT Paper, no. 96. Trier, Ger- many: Linguistic Agency of the University of Trier. Landsberg, Marge E., ed. 1995. Syntactic iconicity and linguistic freezes: The human dimension. Berlin: Mouton de Gruyter. Langacker, Ronald W. 1985. Observations and speculations on subjectivity. In John Hai- man, ed., Iconicity in syntax 109–50. Amsterdam: John Benjamins. Langacker, Ronald W. 1987. Foundations of cognitive grammar. Vol. 1, Theoretical prereq- uisites. Stanford, CA: Stanford University Press. Langacker, Ronald W. 1990. Concept, image, and symbol: The cognitive basis of grammar. Berlin: Mouton de Gruyter. Langacker, Ronald W. 1991. Foundations of cognitive grammar. Vol. 2, Descriptive appli- cation. Stanford, CA: Stanford University Press. Mayerthaler, Willy. 1980. Ikonismus in der Morphologie. Zeitschrift fu ¨ r Semiotik 2: 19–37. Mayerthaler, Willy. 1988 . Morphological naturalness. Ann Arbor, MI: Karoma. Motivation et iconicite ´ . 1993. Special issue of Faits de Langues 1. Newmeyer, Frederick J. 1992. Iconicity and generative grammar. Language 68: 756–96. Peirce, Charles Sanders. [1931] 1974. Collected papers of Charles Sanders Peirce. Ed. Charles Hartshorne and Paul Weiss. Cambridge, MA: Harvard University Press. Pharies, David A. 1985. Charles S. Peirce and the linguistic sign. Amsterdam: John Benjamins. Posner, Roland. 1986. Iconicity in syntax: The natural order of attributes. In Paul Bouissac, Michael Herzfeld, and Roland Posner, eds., Iconicity: Essays on the nature of culture: Festschrift for Thomas A. Sebeok on his 65th birthday 305–37.Tu ¨ bingen: Stauffenburg. Rijkhoff, Jan. 1992. The noun phrase: A typological study of its form and structure. Am- sterdam: John Benjamins. Ross, John. 1980. Ikonismus in der Phraseologie. Zeitschrift f € ur Semiotik 2: 39–56. Ruwet, Nicolas. 1984. Je veux partir/*Je veux que je parte: A propos de la distribution des comple ´ tives a ` temps fini et des comple ´ ments a ` l’infinitif en franc¸ais. Cahiers de Grammaire 7: 76–138. Saussure, Ferdinand de. [1916] 1967. Cours de linguistique ge ´ ne ´ rale. Paris: Payot. iconicity 417 Schachter, Paul. 1974. A non-transformational account of serial verbs. Studies in African Linguistics Supplement V: 253–69. Seiler, Hansjakob. 1978. Determination: A functional dimension for inter-language com- parison. In Hansjakob Seiler, ed., Language Universals 301–28.Tu ¨ bingen: G. Narr. Simone, Raffaele, ed. 1994. Iconicity in language. Amsterdam: John Benjamins. Song, Jae Jung. 1992. A note on iconicity in causatives. Folia Linguistica 26: 333–38. Stassen, Leon. 1985. Comparison and Universal Grammar: An essay in Universal Grammar. Oxford: Basil Blackwell. Swiggers, Pierre. 1993. Iconicite ´ : Un coup d’oeil historiographique et me ´ thodologique. Faits de langue 1: 21–28. (Special issue on ‘Motivation et Iconite ´ ’) Thompson, Sandra A. 1995. The iconicity of ‘dative shift’ in English: Considerations from information flow in discourse. In Marge E. Landsberg, ed., Syntactic iconicity and linguistic freezes: The human dimension 155–75. Berlin: Mouton de Gruyter. Tomlin, Russell S. 1986. Basic word order: Functional principles. London: Croom Helm. Van Langendonck, Willy. 1979. Definiteness as an unmarked category. Linguistische Ber- ichte 63: 33–55. Van Langendonck, Willy. 1995. Categories of word order iconicity. In Marge E. Landsberg, ed., Syntactic iconicity and linguistic freezes: The human dimension 79–90. Berlin: Mouton de Gruyter. Van Langendonck, Willy. 1999. Markedness and prototypical speaker attributes. In Leon de Stadler and Christoph Eyrich, eds., Issues in cognitive linguistics 1993: Proceedings of the 3rd international cognitive linguistics conference 567–76. Berlin: Mouton de Gruyter. Van Langendonck, Willy. 2004. Proper names and forms of iconicity. Logos and Languages: Journal of General Linguistics and Language Theory 5: 15–30. (Special issue, Syntactic categories and parts of speech, ed. Klaas Willems) Van Langendonck, Willy. 2007. Theory and typology of proper names. Berlin: Mouton de Gruyter. Van Langendonck, Willy, and Wim de Pater. 1993. Ikonizita ¨ t in natu ¨ rlicher Sprache. Kodikas 15: 1–20. Van Langendonck, Willy, Pierre Swiggers, and Mark Van de Velde. Forthcoming. The relator- principle as an explanatory parameter in linguistic typology: An exploratory study of comparative constructions. In Peter Lauwers and Pierre Swiggers, eds., Linguistic concepts and currents. Leuven, Belgium: Peeters. Van Langendonck, Willy, Pierre Swiggers, and Mark Van de Velde. N.d. The North Eu- ropean belt of RIM languages. Manuscript. Wierzbicka, Anna. 1975. Why ‘kill’ does not mean ‘cause to die’: The semantics of action sentences. Foundations of Language 13: 491–528. Zipf, George K. [1949] 1965. Human behavior and the principle of least effort: An introduction to human ecology. New York: Hafner. 418 willy van langendonck part ii MODELS OF GRAMMAR . verbs, the first relatum functions as the subject of the iconicity 411 verb. With nonpredicative relators, the first relatum is either the relator’s head or the subject of the predicate of a clause Bill, the subject John refers to the agent of the action of killing, of which the patient, expressed by the object Bill, is the victim. We take predicative relators to refer here to all sorts of. Alice. The first relatum is either the relator’s head, as in (29a)–(29c): girl, excitement, did, or the subject of the clause’s predicate, as in the comparative sentence (29d): Kevin. (29) a. the