A LEXANDER B ERGS When concepts clash ... 1 Coercion, in a nutshell, is what we can find when two linguistic elements which are actually incompatible are combined. This could either result in a completely unacceptable combination (*I the read a book), or the combination may somehow be acceptable and may make sense as in “I have had too much book today” or “I had three coffees this morning”. Much is a modifier for uncountables, book is countable, a combination of the two should not be OK. Still this sentence (I have had too much book today) can be parsed. And the same applies to three coffees. The numeral three suggests countability, coffee is an uncountable noun, and a combination of the two should not be acceptable. However, constructions like this are actually in use. Similar examples include sentences like: “You have apple on your shirt”. Apple is countable and should not be available here as a kind of substance, as in, for example, “you have gravy on your shirt”. And again, while the sentence as such should actually not be acceptable, occurrences like these are not uncommon. In fact, there are many more examples that would probably not be found in a standard grammar book and that would yet be perfectly acceptable in everyday spoken English. In many of these cases, speaker-hearers arrive at an interpretation of the utterance through coercion, i.e. through a reinterpretation that one of the items or the construction as such forces onto the other. Ziegeler (992) defines coercion simply as the resolution of mismatch. Note that coercion is actually invisible in the item itself and in the syntactic environment, i.e. it is rather unpredictable. Coercion usually takes place when specific items are used in certain specific linguistic and/ or extra-linguistic contexts. Three coffees is great, three teas is a little bit more awkward, but probably still OK, three waters is, arguably, very strange and not really acceptable. All three, coffee, tea, water are regular, liquid uncountables, and they are put into one and the same environment. I think it is quite clear that there is hardly any predictable grammatical factor here (and I leave it up to the reader to test other liquids, from beer through milk to tequila). The situation is even further complicated by the fact extralinguistic factors seem to influence the acceptability of potentially coercive combinations. While three coffees is fine in almost any social context today, three waters is only (or at least much more) acceptable in restaurant scenes or in scenes where water is regularly discussed or served. The point is that coercion in linguistics seems to be closely connected to both linguistic (co-textual) and non-linguistic (contex- 1 This paper is dedicated to my son Julius, who spent much of the time when it was written on my lap, and apparently provided excellent inspiration. A LEXANDER B ERGS 62 tual) factors. Only in certain environments and situations do hearers have the chance to see the interpretation intended by the speaker. So when these ‘awkward’ constructions are used they can trigger „the need to resolve semantic conflict in contexts which require particular reinterpretation mechanisms to be fully processed“ (Ziegeler 992). Many approaches to coercion distinguish between what has been called implicit and explicit type-shifting. Type-shifting here simply means that „a construction denotes a different kind of entity or event from the lexical expression with which it is combined“ (Michaelis 2004: 29). Explicit typeshifting occurs when a lexical item (or its projection) is shifted in its designation by the grammatical construction with which it is usually combined. One textbook example is the mismatch between uncountable mass nouns like bread and partitive expressions like a piece of, which essentially refer to countable portions. Obviously the two can be combined successfully into a piece of bread. This is what is called explicit type-shifting, since the shift is directly derivable from the combination of parts. It is predictable in some way. Things are very different when we have a countable noun such as sheet. This can also combine with a partitive construction, but here a new meaning is contextually forced into the lexical item: “give me a shred of sheet”. Here, sheet must be rather seen like an uncountable substance. This has been called implicit type-shifting, since it is neither predictable nor derivable once we know the parts that combine. And there is one more concept needed, the socalled override principle. The override principle basically says that the meaning of lexical items is adjusted to, or overridden by, the syntactic configurations in which they occur. In other words, syntactic patterns essentially determine meaning, not just single lexical items. It is worth pointing out here that coercion may sound weird and rare and awkward, but that many scholars see coercive phenomena even in everyday grammatical structures. There is nominal coercion as in “She had a beer”, where beer is made countable, aspectual coercion as in “I’m loving it! ”, where to love is made dynamic, complement coercion as in “linguistics bores me”, where linguistics is made animate, and quite a few more. So it seems indeed fair to say that coercion as such is part and parcel of everyday communication, that it is the rule rather than the exception. The question now is, however: what does that have to do with the literary mind? First of all there is good reason to believe that coercion and its basic mechanisms are not just some technical gimmick that linguists came up with to make the data fit their theories. Rather, it can be assumed that coercion, and the resolution of mismatch are very central, cognitively grounded and well-motivated mechanisms which are independent of language as such. This is reflected partly in the pragmatic slogan coined by the psychologist Bartlett in 1932: “effort after meaning” (20, 44). Apparently, human beings invest a remarkable amount of effort and energy to make sense of things they perceive, to find, or create meaning, even when the things they encounter super- When concepts clash ... 63 ficially are incompatible and do not “make sense”. So when we do encounter apparent nonsense we have an inbuilt drive (or need) to make sense of it, to find some meaning. In linguistic pragmatics, for example, we have the wellknown cooperative principle developed by Grice with its four maxims of Quantity, Quality, Relevance, and Manner. These maxims are the foundation of successful communication: speak as much as is required (quality), only say what you believe to be true (quality), be relevant (relevance), and be orderly and clear (manner). However, in our daily communication we do not always follow the maxims. In fact, we flout them quite often. The flouting of conversational maxims, however, does not lead to a breakdown in communication, but rather helps addressees to see that the speaker probably performed an indirect speech act which now needs to be recovered. Quite simply: when somebody in a faculty meeting exclaims, out of the blue, “It’s cold in here! ”, he or she certainly does not want to describe the room temperature. The utterance thus flouts the maxim of relevance. But this obvious violation leads the hearers to the suspicion (on the basis of ‘effort after meaning’) that something else must have been intended: “Close the window”, “Turn up the heating”. If all this is true, and coercion plus effort-after-meaning are more general, cognitive phenomena, it seems quite plausible to assume that they are also operational in other domains, and not just everyday language. Or if we want to phrase that as a question: does the literary mind also work with coercion? In the following, I will argue that it does indeed. It might even be argued that mismatch and coercion are an integral part of many literary works, and that they are relevant on a number of different levels. In the following, I will discuss several types of literary products and literary effects, starting with poetry and the lexicon-syntax interface. Quite a number of literary strategies, first and foremost in poetry, operate with what we would now have to call mismatch. Some examples: Death, be not proud, thou some have callèd thee Mighty and dreadful, for thou art not so; (Donne 1296) Quite obviously, this is a case of complement coercion: death can’t be proud, i.e. the semantics of the subject and its corresponding predicate are incongruent and should not go together as such. Similarly, E. E. Cummings provides us with excellent material: O sweet spontaneous earth how often have the doting fingers of prurient philosophers pinched and poked A LEXANDER B ERGS 64 thee , has the naughty thumb of science prodded thy beauty . (e.e. cummings 272) Again, this is a case of complement (or maybe modifier) coercion: earth can’t be sweet or spontaneous, at least not in a straightforward way. Also, pinching and poking is not what philosophers’ (doting) fingers usually do and science does not have a thumb. In the first case, the modifying adjectives sweet and spontaneous are usually not available for the head noun earth. In the second case, subject and predicate are incongruent. In the third case, we find a possessive or partitive construction with elements that are incompatible: science is an abstract noun and therefore does not allow for concrete parts to be attached to it. Other literary genres also abound in coercive phenomena, even children’s books are full of semantico-grammatical mismatches. One example is “Wolves stopped being dribbly” (David Melling, The Kiss That Missed). The concept of dribbliness is usually not associated with wolves. So, in sum, it seems that coercion on this level, since it is so much based on daily language use, is also natural and wide-spread in literature. One might even take this one step further and argue that on this general level, coercion and mismatch are very helpful tools for speakers to express their inbuilt linguistic creativity. Keller, Haspelmath, Haiman, and many others have pointed out that speakers usually like to play around with their language, to be new, innovative, expressive with what they say. Speakers simply like to bend the rules of their language. On the level that has just been described this is not only quite easy but also a very effective and powerful expressive device in everyday language and also in literature. But mismatches in literature do not just occur on the simple semanticogrammatical level. Looking at the well-known poem “Anthem for Doomed Youth” by Wilfred Owen, there is another interesting mismatch to be found here: What passing-bells for these who die as cattle? - only the monstrous anger of the guns. Only the stuttering rifles' rapid rattle Can patter out their hasty orisons. No mockeries now for them; no prayers nor bells; Nor any voice of mourning save the choirs, - The shrill, demented choirs of wailing shells; And bugles calling for them from sad shires. What candles may be held to speed them all? Not in the hands of boys but in their eyes When concepts clash ... 65 Shall shine the holy glimmers of goodbyes. The pallor of girls' brows shall be their pall; Their flowers the tenderness of patient minds, And each slow dusk a drawing-down of blinds. (Owen 540f.) Apart from the many mismatches on the level of semantics and grammar (e.g., “anger of the guns”), it is the form of the poem itself which is actually incongruent with its message, its content. The form, obviously, is that of a traditional Shakespearean sonnet, and thus usually associated with love, requited love, beauty, maybe even politics, mortality. Shakespearean sonnets usually do not describe war, the details of warfare and fighting scenes, or their consequences in such a brutal and open way as Owen’s poem. In contrast to many Shakespearean sonnets, it is somehow shockingly unambiguous in its message and function. What we see here might be regarded as an interesting clash of form and function on a higher level. Needless to say, this is not as common in everyday language. Although even here form and content sometimes clash. This can even be taken a little bit further. At least since Sterne’s Tristram Shandy, authors have played around with literary forms, have breached conventions, have stretched limits, have transgressed the boundaries of tradition. Some of the latest examples are Danielewski’s House of Leaves and Only Revolutions. Both these works which superficially may look like novels are nowhere near what Ian Watt, and with him Defoe, Richardson, and Fielding would probably have accepted as such. Similar phenomena in other genres, such as drama, have to be discussed in a separate paper. In sum, the diagnosis is fairly clear: there are numerous instances and cases of mismatches in literature. What remains to be done now is to show in how far these mismatches and their processing are similar to the mismatches and their resolution through coercion in linguistics. For the sake of clarity, let me repeat the main features of mismatches and coercion in linguistics. Explicit type-shifting happens when the combination of two elements predictably leads to new interpretations of one of the two. Implicit type-shifting is more or less unpredictable. The override principle essentially says that other things being equal, syntactic structures win out over lexical insertions, and so determine the meaning of a construction (e.g., bread looses its meaning when it is inserted into the partitive construction). There is reason to argue that metonymy, for example, as in “The ham sandwich wants to pay”, is usually a question of explicit type-shifting since metonymic processes are normally contiguous and, given enough cultural background, fairly easy to decode or predict. Their composition is, on the surface, incompatible, but their meaning is usually derivable from their parts plus contextual background. Note that the latter can be very important. Jackendoff (53) claims that “He poured three cements today“ is not possible, while Ziegeler (1002) points out that in certain context (e.g. on a specialized construction site) the utterance might not be so bad. Metaphor, on the other A LEXANDER B ERGS 66 hand, seems to be an example of implicit type-shifting, since metaphors are usually based on analogy (see Hopper and Traugott), but still unpredictable as such. These two facts already account for quite a few simple, basic mismatch resolutions in literature. It gets much more interesting when we turn to the larger structures. How do people resolve conflicts between form and function on the level of genre and the like? First, it needs to be pointed out of course that they need to perceive the mismatch as such. In contrast to linguistic structures, where explicit type-shifting (like metonymy) might be something much more intuitive, here we need active, intentional participation. Only somebody who is familiar with Shakespearean sonnets will be able to see the dissonance and mismatch in Owen’s poetry. Somebody who does not know what a Shakespearean sonnet is will simply miss this point. N.B., this is not to say that readers without this background cannot understand the work as such. It means quite simply that such readers will miss this particular point in their interpretation. Their reading, however, might highlight a number of other fascinating and equally interesting aspects. But even if you perceive the mismatch, the question certainly is: how do you resolve it? We might find some kind of override principle here, namely that external form (which we usually perceive first, before we look at content) usually determines our expectations and our interpretation. Danielewski’s House of Leaves, Sterne’s Tristram Shandy, Joyce’s Finnegan’s Wake essentially count as novels and are treated as such. They have the shape, size, and prize of novels, are marketed as such and would be sold as novels. Hence the potential readers would approach these works as novels, i.e. as works of fiction, which essentially follow the rules that define this particular genre for us today. What about mismatches in content, then? What about the postmodern novel, such as Rushdie’s Midnight’s Children, Satanic Verses, or perhaps even Nick Hornby’s High Fidelity? Do readers discard them as nonsensical? Obviously they don’t. Do they create a new category for them? They don’t. Again, they count as novels with a narrative structure. And readers try to find or create meaning on that basis. When conventions are broken, people go out hunting for meaning. And this is also the main reason why we have mismatches and coercion in the first place. We are homines ludenses, and we like to play around with essentially everything we find, including language, linguistic products. Playing around with language, being witty and innovative provides us with evolutionary advantages, since it keeps us and others entertained (see Keller, Haspelmath), playing around with literary form, from the very basic level up to most complex questions of composition and genre helps authors to generate meaning. They can rely on the fact that, a cooperative principle of some sort and the effortafter-meaning-principle will prod to find meaning behind seemingly incongruous constructions. Readers can trust that authors actually tried to make sense somehow, that there is something to find or create. This creation of and search for meaning, and the use of mismatch and coercion are something deeply engrained in our linguistic and literary minds. When concepts clash ... 67 Works Cited Bartlett. Remembering. Cambridge: Cambridge University Press, 1932. Cummings, E.E. “O Sweet“. Englische und Amerikaniche Dichtung 4. Ed. Eva Hesse and Heinz Ickstadt. München: C.H. Beck, 2000. Donne, John. “Death be not proud”. 1633. The Norton Anthology of English Literature. 8th ed. New York; London: Norton, 2006. 1296. Haiman, John. Talk Is Cheap. Sarcasm, alienation, and the evolution of language. Oxford: Oxford University Press, 1998. Haspelmath, Martin. “Why is grammaticalization irreversible? “ Linguistics 37.6 (1999): 1043-1068. Hopper, Paul, and Elizabeth Traugott. Grammaticalization. Cambridge: Cambridge University Press, 2006. Jackendoff, Ray. The architecture of the language faculty. Cambridge, MA: MIT Press, 1997. Keller, Rudi. Sprachwandel: Von der unsichtbaren Hand in der Sprache. Tübingen: Francke, 1990. Melling, David. The Kiss That Missed. London: Hodder Children’s Books, 2003. Michaelis, Laura. “Type Shifting in Construction Grammar: An Integrated Approach to Aspectual Coercion“. Cognitive Linguistics 15.1 (2004.): 1-67. Owen, Wilfred. “Anthem for a Doomed Youth”. 1917. The Norton Anthology of of Modern Poetry. 2nd ed. New York; London: Norton, 1988. 540f. Watt, Ian. The Rise of the Novel. London: Chatto and Windus, 1957. Ziegeler, Debra. “A word of caution on coercion“. Journal of Pragmatics 39.5 (2007): 990- 1028.