inferable from the context (factor Ga1), though not specifically represented by the linguistic forms themselves. Agency is slightly more salient in (31b), where the intransitive verb pass includes indirect reference to an agent within its lexicaliza- tion (factor Ab3). Still more attention is on agency in (31c), whose passive syntax (in construction with a now transitive verb pass) directly represents the presence of an agent (factor Ba 4—not included above). A sharp rise in attention on the agent occurs when it is explicitly referred to by an overt pronoun (factor Ga1), the oblique them in (31d). The agency is further foregrounded by the occurrence of this pro- noun as subject in initial position (factors Ba1 and Ba2)in(31e). And finally, re- placement of the pronoun by a full lexical noun (factor Aa1), as in (31f), fore- grounds the Agent to the greatest degree. (31) a. The goblet slowly went around the banquet table. b. The goblet slowly passed around the banquet table. c. The goblet was slowly passed around the banquet table. d. The goblet was slowly passed around the banquet table by them. e. They slowly passed the goblet around the banquet table. f. The diners slowly passed the goblet around the banquet table. 3.2. Reinforcement of an Attentional Pattern through Factor Convergence Several factors can converge on the same linguistic entity to reinforce a particular level of salience, making it especially high or especially low. The grammar of a lan- guage is often organized so as to facilitate certain convergences. Thus, as seen in the final example sentence of the preceding series, (31f), English regularly foregrounds the concept of agency strongly through the convergence of all the following factor values: explicit representation (Ga1) by an open-class nominal (Aa1) in initial sentence position (Ba1) as grammatical subject (Ba2) of a verb lexicalized to apply to an Agent subject (Ab3). 3.3. Attentional Resultants of Factor Conflict Two factors can conflict in their attentional effects, with the resolution usually either that one factor overrides the other or that they are in competition in which case the hearer’s attention is divided or wavering between the two claims on it. For an example of override, consider the sentence in (32a). Here, the concept of ‘air- craft’ is relatively foregrounded in the constituent plane through the convergence of four factors. It is expressed in the lexical category highest on the attentional hier- archy, a noun (Aa1); it is the sole concept expressed in its morpheme (Ab1); it is in the prominent sentence-final position (Ba1); and it receives the heavy stress stan- dard for such a final constituent (Dc4). By contrast, the same concept of ‘aircraft’ is 290 leonard talmy relatively backgrounded within the constituent flew in (32b). It is backgrounded there through the same four factors: it appears in a lexical category lower on the attentional hierarchy, a verb; it is joined there by other concepts, namely, ‘go’ and ‘by means of ’; it is in a sentence position nonprominent in English, and it receives the relatively low stress of that position. Accordingly, an English speaker may hear this latter sentence as mainly conveying the fact of the journey per se to Key West and as including the idea of aeronautic means only as incidental background in- formation. However, the further application of extra heavy stress (factor Db1)to the verb, as in (32c), now undoes the backgrounding effects of the four convergent factors. It overrides them and forces the foregrounding of the ‘aircraft’ concept. (32) a. I went to Key West last month by plane. b. I flew to Key West last month. c. I FLEW to Key West last month. In the competition type of conflict, each of two or more factors calls on the hearer’s limited attentional capacity for its own target, with the consequence that one or more of the targets receives less attention than it needs for adequate pro- cessing. For example, factor Ha1 calls on the hearer to allocate enough attention to the speaker’s currently uttered forms for them to be processed in working memory. But if the speaker had just previously uttered an ill-formed sentence, factor F3 calls on the hearer to allocate enough attention to the discrepancy to puzzle out what the speaker might have intended to say. The hearer may not have enough atten- tional capacity to act on both factors adequately at the same time. The hearer might attend to the current words and leave the earlier undecipherable discourse unre- solved, or he or she may work on the prior discourse while missing what is now being said, or even may allocate some attention to each task, performing neither of them well. REFERENCES Asher, Ron E. 1985. Tamil. Beckenham, UK: Croom Helm. Bierwisch, Manfred. 1983. Psychologische Aspekte der Semantik natu ¨ rlicher Sprachen. In Wolfgang Motsch and Dieter Viehweger, eds., Richtungen der modernen Semantik- forschung 15–64. Berlin: Akademie Verlag. Brugman, Claudia, and George Lakoff. 1988. Cognitive topology and lexical networks. In Steven L. Small, Garrison W. Cottrell, and Michael K. Tanenhaus, eds., Lexical am- biguity resolution: Perspectives from psycholinguistics, neuropsychology, and artificial intelligence 477–508. San Mateo, CA: Morgan Kaufmann. Chafe, Wallace. 1994. Discourse, consciousness, and time: The flow and displacement of conscious experience in speaking and writing. Chicago: University of Chicago Press. Fillmore, Charles J. 1975. An alternative to checklist theories of meaning. Berkeley Lin- guistics Society 1: 123–31. attention phenomena 291 Fillmore, Charles J. 1976. Frame semantics and the nature of language. Annals of the New York Academy of Sciences: Conference on the Origin and Development of Language and Speech 280: 20–32. Fillmore, Charles J. 1982. Frame semantics. In Linguistic Society of Korea, ed., Linguistics in the morning calm 111–37. Seoul: Hanshin Publishing. Fillmore, Charles J. 1997. Lectures on deixis. Stanford, CA: CSLI Publications. (First pub- lished as Santa Cruz Lectures on Deixis. Bloomington: Indiana University Linguistics Club, 1975) Givo ´ n, Talmy. 1990. Syntax: A functional-typological introduction. Vol. 2. Amsterdam: John Benjamins. Husserl, Edmund. 1970. Logical investigations. Trans. J. N. Findlay. 2 vols. London: Rou- tledge and Kegan Paul. Lakoff, George. 1987. Women, fire and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press. Lambrecht, Knud. 1994. Information structure and sentence form: A theory of topic, focus, and the mental representations of discourse referents. Cambridge: Cambridge University Press. Langacker, Ronald W. 1984. Active zones. Berkeley Linguistics Society 10: 172–88. Langacker, Ronald W. 1987. Foundations of cognitive grammar. Vol. 1, Theoretical pre- requisites. Stanford, CA: Stanford University Press. Liddell, Scott K. 2003. Grammar, gesture, and meaning in American Sign Language. Cam- bridge: Cambridge University Press. Talmy, Leonard. 1972. Semantic structures in English and Atsugewi. PhD dissertation, University of California at Berkeley. Talmy, Leonard. 1976. Semantic causative types. In Masayoshi Shibatani, ed., Syntax and semantics, vol. 6, The grammar of causative constructions 43–116. New York: Academic Press. Talmy, Leonard. 1978a. Figure and ground in complex sentences. In Joseph Greenberg, ed., Universals of human language, vol. 4, Syntax 625–49. Stanford, CA: Stanford University Press. Talmy, Leonard. 1978b. Relations between subordination and coordination. In Joseph H. Greenberg, ed., Universals of human language, vol. 4, Syntax 487–513. Stanford, CA: Stanford University Press. Talmy, Leonard. 1978c. The relation of grammar to cognition—a synopsis. In David Waltz ed., Proceedings of TINLAP-2 14–24. New York: Association for Computing Machinery. Talmy, Leonard. 1983. How language structures space. In Herbert L. Pick, Jr., and Linda P. Acredolo, eds., Spatial orientation: Theory, research, and application 225–82. New York: Plenum Press. Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In Timothy Shopen, ed., Language typology and syntactic description, vol. 3, Grammatical categories and the lexicon 57–149. Cambridge: Cambridge University Press. Talmy, Leonard. 1988a. Force dynamics in language and cognition. Cognitive Science 12: 49–100. Talmy, Leonard. 1988b. The relation of grammar to cognition. In Brygida Rudzka-Ostyn, ed., Topics in cognitive linguistics 165–205. Amsterdam: John Benjamins. Talmy, Leonard. 1991. Path to realization: A typology of event conflation. Berkeley Lin- guistics Society 17: 480–519. Talmy, Leonard. 1995a. The cognitive culture system. The Monist 78: 80–116. 292 leonard talmy Talmy, Leonard. 1995b. Narrative structure in a cognitive framework. In Gail Bruder, Judy Duchan, and Lynne Hewitt, eds., Deixis in narrative: A cognitive science perspective 421– 60. Hillsdale, NJ: Lawrence Erlbaum. Talmy, Leonard. 1996a. Fictive motion in language and ‘‘ception.’’ In Paul Bloom, Mary Peterson, Lynn Nadel, and Merrill Garrett, eds., Language and space 211–76. Cam- bridge, MA: MIT Press. Talmy, Leonard. 1996b. The windowing of attention in language. In Masayoshi Shibatani and Sandra Thompson, eds., Grammatical constructions: Their form and meaning 235–87. Oxford: Oxford University Press. Talmy, Leonard. 2000a. Toward a cognitive semantics. Vol. 1, Concept structuring systems. Cambridge: MIT Press. Talmy, Leonard. 2000b. Toward a cognitive semantics. Vol. 2, Typology and process in concept structuring. Cambridge: MIT Press. Talmy, Leonard. 2003a. Recombinance in the evolution of language. Chicago Linguistic Society 39 (The Panels). Talmy, Leonard. 2003b. The representation of spatial structure in spoken and signed language. In Karen Emmorey, ed., Perspectives on classifier constructions in sign lan- guage 169–95. Mahwah, NJ: Lawrence Erlbaum. Talmy, Leonard. Forthcoming. How language directs attention. Cambridge: MIT Press. Tomlin, Russell S. 1995. Focal attention, voice, and word order: An experimental, cross- linguistics study. In Pamela Downing and Michael Noonan, eds., Word order in discourse 517–54. Amsterdam: John Benjamins. Zubin, David A., and Lynne E. Hewitt 1995. The deictic center: A theory of deixis in narrative. In Judith F. Duchan, Gail A. Bruder, and Lynne E. Hewitt, eds., Deixis in narrative: A cognitive science perspective 129–55. Mahwah, NJ: Lawrence Erlbaum. attention phenomena 293 chapter 12 FORCE DYNAMICS walter de mulder 1. Introduction According to Leonard Talmy (2000: 4), Cognitive Semantics is the study of the way conceptual content is organized in language. In Talmy’s view, a sentence (or other portion of discourse) does not objectively represent its referent scene, but it evokes in the listener a cognitive representation, defined as ‘‘an emergent, compounded by various cognitive processes out of the referential meanings of the sentence elements, understanding of the present situation, general knowledge, and so on’’ (2000: 93, note 2). Talmy furthermore proposes that the grammatical, closed-class elements in a sentence convey the structure of the cognitive representation, while the lexi- cal, open-class elements mainly contribute to sentence content. Across languages, the set of grammatically specified notions ‘‘collectively constitutes the conceptual structuring system of language’’ (2000: 21), which patterns in five schematic systems (formerly called ‘‘imaging systems,’’ see Talmy 2000: 40): a. ‘‘Configurational structure’’ imposes a particular spatial and temporal structure on referent scenes viewed as composed of entities of a particular nature in particular relationships (Talmy 2000: 47; 467). b. ‘‘Location of perspective point’’ adds the point of view from which the scene is conceived, or ‘‘ceived’’ (for the notion of ‘‘ceiving,’’ see Talmy 1996a; 2000: 99–175). c. ‘‘Distribution of attention’’ concerns the way attention is distributed over aspects of the referent scene and assigns it a Figure/Ground distinction. d. ‘‘Force dynamics’’ ‘‘pertains to the linguistic representation of force inter- actions and causal relations occurring between certain entities within the structured situation’’ (Talmy 2000: 12). e. ‘‘Cognitive state’’ is concerned with the speaker’s knowledge status or the referent scene’s status of reality, as expressed by mood, evidentials, and modal verbs (Talmy 2000: 1–18, 41, 56, 92). The schematic system ‘‘Force dynamics’’ is first defined by Talmy as a fun- damental semantic category in the realm of physical force and is viewed in par- ticular as a generalization over the linguistic notion of ‘‘causative’’ (see Talmy 1981, 1985, 1988, 2000: 409–70). Metaphorical transfers subsequently generalize force- dynamic conceptions to the domains of internal psychological relationships and social interactions. As such, for instance, the system of English modals is analyzed in force-dynamic terms before it is shown how force dynamics also partially struc- tures discourse and argumentation. I will largely follow this order of presentation in section 2 of this chapter. In section 3, I will turn to some of the research en- gendered by the notion of force dynamics, such as Jackendoff’s (1990) endeavor to incorporate force dynamics in his conceptual semantics, and several studies of modal verbs in Cognitive Linguistics. In the final section, I will discuss how com- parable notions, such as energy transfer, as they are introduced in Langacker (1990, 1991), can be usefully employed in the definition of grammatical categories such as ‘subject’ and ‘object’ and in the analysis of syntactic structure. 2. Talmy’s View on Force Dynamics as a Fundamental Linguistic Category 2.1. The Fundamental Patterns A force-dynamic pattern which underlies ‘‘all more complex force-dynamic pat- terns is the steady-state opposition of two forces’’ (Talmy 2000: 413). In language, the two participants in this fundamental scene are assigned two different semantic roles: (i) the first participant, the ‘‘Agonist,’’ is the participant in focal attention, since the salient issue in the interaction is whether the Agonist can manifest its force tendency or not; (ii) the second participant, the ‘‘Antagonist,’’ is considered for the effect it has on the Agonist, namely, overcoming the latter’s force tendency or not. Language presents entities as possessing intrinsic force tendencies: they are held to tend intrinsically toward motion or rest, or rather toward action or inac- tion. Their confrontation yields a resultant, which presents the Agonist as being either in action or in inaction. This linguistic conception of forces and their interaction results in four basic force-dynamic patterns, which are illustrated in sentences (1a)–(1d): 1 force dynamics 295 a. The Agonist’s intrinsic tendency toward rest is overcome by a stronger Antagonist, which forces it to move. b. The Agonist’s tendency toward rest is stronger than the force opposing it; consequently, the Agonist remains in place. c. The Agonist’s inherent tendency toward motion is opposed by the Antagonist, but the Agonist is the stronger entity. d. The Agonist has a tendency toward motion, but the Antagonist is stronger and blocks it. (1) a. The ball kept rolling because of the wind blowing on it. b. The shed kept standing despite the gale wind blowing against it. c. The ball kept rolling despite the stiff grass. d. The log kept lying on the incline because of the ridge there. Patterns a and d, illustrated in (1a) and (1d), where the Antagonist continuously impinges on the Agonist, are called cases of ‘‘extended causation’’ (see Talmy 2000: 415–16). Diagrammatically, the four basic force-dynamic patterns are represented in figure 12.1. Adopting ‘‘force dynamics’’ as a semantic category enables us to integrate in a unified frameworknotions expressed bydifferent grammatical categories: ‘‘honorary’’ auxiliaries such as keep -ing (as in 1a–1d); conjunctions such as because (1a, 1d) or although; prepositions such as because of (1a, 1d) and despite (1b, 1c), but also against (see also Schepping 1991 on French contre; Kristoffersen 2001 on Norwegian mot;and Belie ¨ n 2002 on Dutch tegen), German € uber (Meex 2002), or French aa travers ‘through’ Figure 12.1. The basic steady-state force-dynamic pattern (based on Talmy 2000: 415) 296 walter de mulder (Stosic 2002); and, finally, expressions related to associated schematic models, such as straight (Cienki 1998). 2.2. Complex Patterns By adding supplementary factors to the four basic patterns presented above, more complex force-dynamic patterns are created. Change over Time The relation between the Agonist and the Antagonist can change over time, as can be seen in examples ( 2)–(5), where—unlike in (1a)–(1d)—the Antagonist is expressed by the subject. (2) The ball’s hitting it made the lamp topple from the table. (3) The water’s dripping on it made the fire die down. (4) The plug’s coming loose let the water flow from the tank. (5) The stirring rod’s breaking let the particles settle. Unlike examples (1a) and (1d), these sentences exemplify ‘change over time’, in that they do not show steady impingement of the Antagonist on the Agonist; in (2) and (3), the Antagonist enters the state of impingement, while (4) and (5) involve cessation of impingement. Examples (2) and (3) are referred to by Talmy (2000: 418) as cases of ‘‘onset causation’’; furthermore, (2) expresses prototypical causation, as the Antagonist’s action on the Agonist reverses the latter’s intrinsic tendency. Examples (4) and (5), where cessation of impingement is at stake, involve the concept of ‘letting’. Here, the prototypical case is the one where the Antagonist disengages and allows the Agonist to manifest its force tendency (as in 4); in the less prototypical case (5), the Antagonist ceases to impinge on the Agonist and allows it to come to rest. Examples (2)–(5) show that the force-dynamic analysis allows ‘causing’ and ‘letting’ to be grouped together because these concepts are viewed as complexes of more primitive concepts that can recombine in different ways (Talmy 2000: 419, 428). 2 Force dynamics can thus be seen as a generalization over the causatives studied by Talmy (1976; 2000: 471–549). 3 Time can also affect the relation between the Antagonist and the Agonist in a less radical way, when the impingement of the Antagonist on the Agonist does not disappear altogether, but gradually changes the balance of forces between Agonist and Antagonist, as shown in (6): (6) The enemy overcame us as we stood on the border. Finally, corresponding to examples (4) and (5), where the Antagonist moves away from the Agonist, there also exist patterns of ‘‘extended letting,’’ where the Antag- onist remains away: force dynamics 297 (7) The plug’s staying loose let the water drain from the tank. (8) The fan’s being broken let the smoke hang still in the chamber. In fact, these patterns can be seen as the negation of the basic steady-state patterns: 4 whereas in the latter patterns, Agonist and Antagonist are engaged in an opposi- tion of forces, in (7) and (8), absence of possible engagement is expressed. This is why the processes expressed in these sentences are called ‘‘secondary steady-state force-dynamic processes’’: they are derived from the basic steady-state patterns. Agency As is pointed out by Talmy (1976; 2000: 509–14), including an agent in a causative sentence makes it semantically more complex, because the intention of the agent to make something happen initiates a sequence of causally related events, from the volitional act of the agent via the moving parts of his or her body and other in- termediate events to the final event (see also Croft 1991: 163–65). The resulting action sequence can be expressed by mentioning only the agent and the final event, possibly adding the penultimate event, or just its instrument (Talmy 2000: 421): (9) I broke the vase [(by hitting it) with a ball]. The agent can be omitted, as in (10): (10) The ball’s hitting it broke the vase. 2.3. Alternatives of Foregrounding While ‘‘all of the interrelated factors in any force-dynamic pattern are necessarily co-present wherever that pattern is involved’’ (Talmy 2000: 422), they do not all attract the same amount of attention: elements explicitly referred to, expressed earlier in the sentence, or figuring higher in some case hierarchy, tend to receive more attention. The other elements are still implicitly present, though. In (11), for instance, the Agonist is foregrounded as a result of its subject status, whereas the Antagonist is backgrounded, since it is omitted (11a) or expressed as an oblique object (11b); in (12), the distribution of attention is reversed, since now the Antag- onist is in subject position and the Agonist in direct object position: (11) a. The ball kept rolling. b. The ball is rolling because of the wind. (12) a. The wind kept the ball rolling. b. The wind is making the ball roll. 2.4. Metaphorical Extensions Talmy’s analysis of the basic and complex force-dynamic patterns shows that force dynamics is a fundamental category of language. As such, it has served as a basic domain for metaphorical transfers, allowing language users to conceptualize force-dynamic patterns in domains other than the physical one. 298 walter de mulder Force Dynamics in the Psychological Domain Utterances such as (13)–(15) present the self as divided into a desiring part (the Agonist) and a blocking part (the Antagonist), which represents a sense of respon- sibility or propriety and can be regarded as an internalization of social values: (13) He held himself back from responding. (14) He exerted himself in pressing against the jammed door. These opposing elements are also present in (15), although the lexical semantics of the verb refrain has the effect of presenting the psyche as a whole: (15) He refrained from responding. As a consequence of the conception of the self as an entity in which opposing forces are at work, the overt manifestation of force by sentient beings is generally interpreted as arising from psychological driving forces, rather than as originating in the body itself, as can be shown by comparing examples (16) and (17): (16) The new dam resisted the pressure of the water against it. (17) The man resisted the pressure of the crowd against him. Whereas the dam is supposed to stay in place as a result of its physical properties, it is not his physical, bodily properties that make the man resist the pressure of the crowd; rather, it is his underlying psychological force dynamics, in particular a continuous expenditure of force, originating in the goal-oriented part of his psyche. In general, the central part of the psyche is conceived as having a natural ten- dency toward rest, which has to be overcome by an expenditure of energy by some more peripheral part; moreover, the body is viewed as an essentially inert entity, requiring animation to be moved into action. These aspects can be combined as in example (18), which thus illustrates the ‘‘generative’’ capacity of force-dynamic patterns to embed and form increasingly complex patterns: (18) Fear kept preventing the acrobat from letting the elephant hold up his tightrope. Force Dynamics in the Social Domain The force-dynamic pattern can also be used to structure the social domain, as illustrated in (19): (19) a. He’s under a lot of pressure to keep silent. b. Our government exerted pressure on that country to toe our line. c. Getting job security relieved the pressure on her to perform. d. The gang pushed him to do things he didn’t want to do. The metaphorical transfer underlying the examples in (19) is based on the analogy between the direct exertion of force of one object on another in order to make it move or manifest some other action, and ‘‘one sentient entity’s production of stimuli, force dynamics 299 . (17): (16) The new dam resisted the pressure of the water against it. (17) The man resisted the pressure of the crowd against him. Whereas the dam is supposed to stay in place as a result of its. make the man resist the pressure of the crowd; rather, it is his underlying psychological force dynamics, in particular a continuous expenditure of force, originating in the goal-oriented part of. evokes in the listener a cognitive representation, defined as ‘‘an emergent, compounded by various cognitive processes out of the referential meanings of the sentence elements, understanding of the