Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Familiarity, context, and the distinction between literal and nonliteral language Geiger, Odeis 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_fall_phd_geiger_odeis.pdf [ 10.25MB ]
Metadata
JSON: 831-1.0098794.json
JSON-LD: 831-1.0098794-ld.json
RDF/XML (Pretty): 831-1.0098794-rdf.xml
RDF/JSON: 831-1.0098794-rdf.json
Turtle: 831-1.0098794-turtle.txt
N-Triples: 831-1.0098794-rdf-ntriples.txt
Original Record: 831-1.0098794-source.json
Full Text
831-1.0098794-fulltext.txt
Citation
831-1.0098794.ris

Full Text

FAMILIARITY, CONTEXT, AND THE DISTINCTION BETWEEN LITERAL AND NONLITERAL LANGUAGE by ODEIS GEIGER B.Sc., The University of Oregon, 1965 M.A., The University of British Columbia, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Psychology  We accept this thesis as conforming  THE UNIV SITY OF BRITISH COLUMBIA August, 1993 © Odeis Geiger, 1993  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  eA...--(v  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  ^/'9 3  ABSTRACT  Six experiments investigating the relationship of familiarity and context to the processing of metaphor are presented. Experiment 1 compares time to understand sets containing idiom or metaphor targets following 1-2 sentence contexts in four conditions: NONLITERAL, where the context was completed by an idiom or metaphor, PARAPHRASE, where the same context was completed by a literal target with the same meaning, SURPRISE, where the context was completed by a less-anticipated literal target, and LITERAL, where the target from the NONLITERAL condition was used in its literal sense in a different context. In Experiment 2, sets of dead metaphors replaced the idiom sets, and the metaphor sets had novel metaphors. Experiment 3 matched the targets in each condition for overall printed word frequency, to investigate whether word familiarity was interacting with type of usage. It also included an UNFAMILIAR condition, where the same context was completed by a much less familiar word used literally. Experiment 4 took 20 contexts from Experiment 3 and asked subjects to generate their own endings. Experiment 5 replicated Experiment 3 but with a two-target semantic choice instead of a single response.^Experiment 6 shortened contexts and reduced their information content. Its purpose was to see how much context was contributing to understanding, and whether some conditions would be more affected than others. The results may be interpreted as indicating that familiarity with the use of a word is important in determining  -speed  f^D ea^  metaphors could be understood just as quickly as words used literally, but novel metaphors took longer. Contextual expectations are also a powerful adjunct to the understanding process. When expectations are thwarted, errors and understanding time increases. Metaphor understanding is interpreted as a class-inclusion process in the manner described by Glucksberg and Keysar (1990), where a word used metaphorically is viewed as a prototypical exemplar of a hierarchically superordinate class that becomes extended to incorporate the context topic. This process takes time, but metaphors have a response latency advantage over surprising or unfamiliar literal words encountered in context. When context is reduced, metaphors are still advantageous in terms of time, but are less useful to depth of understanding.  iv.  TABLE OF CONTENTS ABSTRACT ^  ii  TABLE OF CONTENTS ^  iv  LIST OF TABLES ^ LIST OF FIGURES ^ LANGUAGE AND MEANING ^  vi 1  THEORIES OF NONLITERAL LANGUAGE COMPREHENSION ^ 16 INTRODUCTION TO THE EXPERIMENTS ^  32  EXPERIMENT 1: PROCESSING IDIOMS AND METAPHORS ^ 41 EXPERIMENT 2: PROCESSING DEAD METAPHORS AND METAPHORS ^ 61 EXPERIMENT 3: THE EFFECTS OF CONVENTIONALIZATION ^ 72 EXPERIMENT 4: VOLITIONAL METAPHOR GENERATION ^ 85 EXPERIMENT 5: SEMANTIC DECISION ^ EXPERIMENT 6: REDUCED CONTEXT ^  94 108  BETWEEN GROUPS ANALYSIS OF EXPERIMENTS 5 & 6 ^ 125 GENERAL DISCUSSION ^  133  CONCLUSION ^  148  REFERENCES ^  149  APPENDICES ^  156  APPENDIX A - EXPERIMENT 1 ^  157  APPENDIX B - EXPERIMENT 2 ^  170  APPENDIX C - EXPERIMENT 3 ^  182  APPENDIX D - EXPERIMENT 4 ^  196  APPENDIX E - EXPERIMENT 5 ^  226  APPENDIX F - EXPERIMENT 6 ^  247  V  LIST OF TABLES Table 1 ^  52  Condition means and standard errors for five experiments. Table 2 ^ Error and time violations for Experiments 5 & 6.  104  .  v i. LIST OF FIGURES Figure 1 ^  44  Sample idiom item from Experiment 1. Figure 2 ^  54  Results of Experiment 1 - Idiom and metaphor portions. (N)nonliteral [idiom/metaphor] (P)paraphrase (L)literal (S)surprise. 65  Figure 3 ^ Sample dead metaphor item from Experiment 2. Figure 4 ^  69  Results of Experiment 2 - Dead and novel metaphor portions. (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise. Figure 5 ^  77  Sample novel metaphor item from Experiment 3. Figure 6 ^  79  Results of Experiment 3 - Dead and novel metaphor portions matched for word frequency. (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise (U)unfamiliar. Figure 7 ^  89  Sample item from Experiment 4, with responses. Figure 8 ^  91  Results of Experiment 4 - Comparison of the printed word frequency of endings generated by only one subject in a _context_ with the—numbsr TehbYdted by more than one subject. —  vii. Figure 9 ^  92  Results of Experiment 4 - Comparison of mean number of subjects making the same response for first generated and last generated words. Figure 10 ^  101  Results of Experiment 5 - Dead and novel portions in a semantic decision task. (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise (U)unfamiliar. Figure 11 ^  113  Sample novel metaphor item form Experiment 5 and its counterpart from Experiment 6. Figure 12 ^  115  Results of Experiment 6 - Dead and novel portions in a semantic decision task with reduced context. (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise (U) unfamiliar. Figure 13 ^  128  Results of Experiments 5 & 6 together - Dead and novel portions in a semantic decision task (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise (U)unfamiliar. Figure 14 ^  129  Error means for Experiments 5 & 6 together - Mean number of errors per subject for dead and novel portions in a semantic decision task. (N)nonliteral (P)paraphrase (L)literal  (s)qurprice (U)unfamiliar.  Metaphor Comprehension -^1  LANGUAGE AND MEANING The question of how human utterances come to contain and convey meaning presents psychology with one of its most intriguing challenges. How is it that we can use language to encode our perceptions, or to recreate others' experiences and beliefs in our own terms? How is it that every isolated human cultural group invented communication systems based on similar principles of named objects and actions, and simultaneously developed the syntactic structures necessary to make the combination of these elements meaningful and efficient? How is it that we can recall, store, and understand the meaning of so many individual words so quickly, even if they are used in novel or ambiguous ways? How is it that any normal child can acquire and use such a complex system so effortlessly? In the study of human cognition, psychology has the potential to investigate and illuminate these fascinating issues. The everyday reality of language masks what an extraordinary process it is. It seems simple to say that the words we use have meanings attached to them; that these meanings are consistent within and between individuals in a culture; and that it is through this quality of language to permit the sharing of experiences in speech and writing that culture and civilization have^  ---i-e-t-wherre)irie what meaning is, problems  Metaphor Comprehension -^2 immediately arise. Most English speaking people have no trouble understanding and using a word like "cute" for example, and yet many of us would be hard pressed to explain exactly what we are trying to convey when we use it, or to explicate the commonality that would make it an equally appropriate description of a riddle, a teenager, the cut of a dress, a ladybug, a melody, and Kevin Costner's behind. While language is the essential attribute that has permitted humans to pass wisdom along from generation to generation and from civilization to civilization, there is some very good evidence that both thought and memory can exist independent of language. The behaviour of pre-linguistic children plainly reveals that their abilities to think and remember are already in use, and their language learning seems to progress from the observation of an object or an event to the desire to acquire a name by which to identify it. Without at least memory and thought, it would be impossible to make use of words once we are provided with them, and Fodor (1975) takes this claim one step further and argues that we must have some kind of internal, representational language innately in place in order for us to make use of the words we learn in our external language of communication. It is indisputably true that we are constantly extracting meaning from nonverbal symbols and events in our environment. We do not first need a name for something in order to think about it, and it can be argued that most insight does not occur to____us--in—t-lie foriro a --  text or  an internal lecture. It  Metaphor Comprehension - ^3 is only when we want to retain, operate upon, or share an idea that encoding it in written or spoken language becomes useful. The work of Heider [Rosch] (1972) has been ingenious in showing that language can be distinct from thought. She discovered that among the Dani, a primitive culture that had only two color words, people were not constrained to perceive the colors in their environment in terms of only those two words, but could make the same distinctions between colors as persons speaking languages with a much wider range of color terms. These findings decisively countered an earlier hypothesis advanced by Whorf (1956), known as the Sapir-Whorf hypothesis, which stated that the categories of our language determine the way our world is perceived and carried the implication that thought and language were synonymous. The Sapir-Whorf hypothesis in its strong form has now largely been discredited among psychologists, but a variant of it emerges in a somewhat different guise in the recent work of George Lakoff (1987; Lakoff & Johnson, 1980). Like Whorf, Lakoff believes that the concepts of a language become part of the structure of thought, and not just its objects. The congruence of these two theories represents a quintessential example of the slipperiness of language. The Whorf-Sapir hypothesis is often identified as the linguistic relativity hypothesis, since it supports the idea that reality is viewed relative to one's linguistic background, and implies that different linguistic groups might c_ode_reality-di fferentIy: - -  -  It is in this way that  Metaphor Comprehension -^4 Lakoff incorporates Whorf. Yet the same hypothesis is also identified with linguistic determinism (Tartter, 1986), since the language one speaks can be said to determine the mental view of reality. This suggests a rigidly categorical view of the meaning of words that seems to be exactly the opposite to what Lakoff is trying to say. If nothing else, such confusions serve as selfreflexive examples in support of a polysemous view of word meaning. Just as Lakoff uses Whorf's deterministic theory of language to represent a relativistic model, he makes his relativistic model represent a deterministic argument: that our concepts arise out of our bodily experience, and that the language we use to express such concepts determines (or at least strongly structures) our thought. Lakoff (1987) labels his theory experiential realism. In his view, linguistic relativity is important because language is embodied, that is, derived from human bodily and/or social or cultural experience. Since an individual can only know his/her own body from the inside, this makes every individual a unique perceiver. Thought and reasoning take the imaginative forms of propositions, metaphor, metonymy, and mental imagery. All concepts are structured, and new ideas are understood when they are incorporated into a conceptual structure based on previous experience. Lakoff calls these conceptual structures Idealized Cognitive Models, or ICMs. ICMs are structured very similarly to the way Rosch (1973) describes for prototype models, as concepts organized • • • • e_ _ -^-cwaities. The ICMs with  Metaphor Comprehension -^5 membership based on family resemblances parallels the way concepts are organized around words. Lakoff bases his belief that thought and language make use of the same cognitive structure on this evidence. One other group of cognitive theories that identifies language with thought are those based on the idea of Parallel Distributed Processing, or PDP (Allman, 1989; Chandler, 1991; Forster, 1990; Hubbell, 1988; McClelland & Kawamoto, 1986). In these models, a specific pattern of neuronal activation (or more accurately, a simulation of a neuronal activation pattern generated by a computer) is distributed over a large matrix of connection points. Information is contained in the weightings of the connections between points, which become strengthened as information is processed. Competing units, meanwhile, become ruled out and inhibited by input from other connections until the network reaches a state of equilibrium. When a connectionist system processes language, the strength of the active connections are all slightly incremented each time a word is recognized. Word frequency effects are an integral part of PDP models. Similar experiences produce consistent activations of neurons, as is generally described for prototype models, and both learning and generalization of concepts can be explained in this way. On the other hand, one limitation of such a model is that since PDP connections are strengthened by repetition, they present a problem for creative or nonliteral thinking, since the likelihood f •en  ing a 'deviant' thought or a new idea  Metaphor Comprehension -^6 would be less probable than repeating an established one. Because it explains activation by semantic similarity, such a model would have a hard time understanding paradoxes or incongruities. In the same vein, it would also appear difficult for a PDP model to account for lies or sarcasm - cases where a person is saying one thing while thinking another. A system that responds to a pattern of activation presumes that activating a specific network of neurons could not deal with two simultaneous and contradictory results. While it may be unfair to characterize them in such a way, PDP models have some common features with the 'language is thought' hypotheses of the Behaviorists, which limit what can be known about thought to what can be observed through language. Each of the above proposals in some way argues in favor of linguistic relativity, or at least that the cognitive activity associated with language processing is inextricably tied to linguistic input. The other side of this coin is the hypothesis that the systems of thought and language are separate, and the strongest proponent of such a hypothesis is Fodor (1975). Fodor argues that thought must have a syntactic structure in the same way as language does, based on the manipulation of symbols in accordance with syntactic rules. He claims that you cannot learn a language whose terms express semantic properties not expressed in a language you are already able to use. Fodor's (1975) 'Language of Thought' model appears to be the least real i stio of all the theories about the relationship -  -  -  -  Metaphor Comprehension -^7 between language and thought. Language is described by a computational theory based upon principles of logic and the combination of representational symbols according to syntactic rules. It demands that words be categorically defined, in order for the correspondence between the innate language to remain consistent with the learned language. Limiting language in this way precludes either linguistic evolution or linguistic flexibility. Most modern cognitive scientists accept the vast body of empirical evidence indicating that human behavior is neither logical nor rule governed, and regard such a theory to be unrepresentative of human cognition. Fodor's (1975) system derives from Chomsky's (1965) belief in linguistic universals, and to be universal, representations must be independent of a specific language. Although both allow that the concepts can be quite primitive, there is still the claim that we somehow innately possess the elements of all that we will ever know. Like Chomsky, Fodor argues that investigation must be limited to the syntactic, computational properties of language, since the semantic component of language is beyond the ability of humans to access or understand. In defense of Fodor and Chomsky, however, is the incontrovertible evidence that all children can learn to use language amazingly quickly and competently, without explicit instruction. It is also true that very young children can become fluently bilingual or trilingual, while such an accomplishment is difficult _and----imper fect far most adults. While it may be going -  -  -  Metaphor Comprehension -^8 too far to argue that we are born with the primitive elements of all the semantic concepts we will encounter in our interaction with the world, it does not seem at all unlikely that at least some syntactic ability must be innate. Questions of whether or not language and thought are identical have profound implications for the representation of language. Dictionaries are familiar structures for organizing words and their meanings within developed cultures, and may be proposed as a likely analogical representation for the way language is stored and accessed in our brains. The problem with a dictionary analog is that it demands a Whorfian conception of thought - that a word and the meaning associated with it can be regarded as a categorical entity, and that a collection of such categorical entities is sufficient to explain how humans use language. The categorical approach to word meaning may be more an artificial tool than a natural capacity. Much current thinking cites the work of Wittgenstein (1953) and Rosch (1973; 1978) with regard to the fact that classical categories consisting of objects and their features do not reflect categories in the real world, where membership can be graded or fuzzy, and where ad hoc categories that cut across otherwise nonintersecting concepts can be assembled at will (i.e., purple things; things you can cut with scissors). Dictionary definitions represent the epitome of classical categorization and suffer the same defects when used as a cognitive model of how human^  ing.  Metaphor Comprehension -^9 In my view, the most interesting and plausible model of deriving meaning from language has its origin in basic level prototype categories. Such a model would represent concrete examples derived from experience and/or learned definitions at a subordinate level and abstract themes at a superordinate level. Category prototypes would often be single words, but could also be word combinations or phrases. Not all the words that are in a dictionary would be represented in prototypes; syntactical words are probably stored and manipulated in a completely different way, and proper nouns may be organized as clusters of distinctive features rather than of similar exemplars. The theorist who I feel presents this model best is Schank (1982; 1990), though there have been many types of schema or script models developed that are all based on similar organizing principles. The major advantage of prototype models is that they explain how individual experiences can be maintained as separate entities, while simultaneously accounting for experience-based learning by extracting common elements from recurrent events. The basis of Schank's theory is that knowledge is organized into Memory Organization Packets (MOPS), which store events, and Thematic Organization Packets (TOPs) which are organized by abstract theme. Schank's MOPs are derived from his earlier scripts, which are short scenarios of the consistent events that make up our life experience. The MOPs are built up very much in the way that Rosch (1973; 1978) describes for prototype models similar, repeated experiences are stored together and become  Metaphor Comprehension - 10 firmly established, while all of the different features of individual episodes, called scenes, remain attached to the part of the script they relate to. The scenes' tendency to become attached to different MOPs presents a plausible model for the way we can flexibly handle ad hoc categories and creative thoughts. Scenes can also function as episodic memories, which we would use to recall a specific instance of the general class of event represented by the MOP. MOPs seem to represent the same level of organization as basic level categories do, and can be events like visiting a business office or washing dishes. MOPs are further organized into TOPs or Thematic Organization Packets. The TOPs are domain-independent themes that the prototypical scenes exemplify. One incident may remind us of another that is not similar in a concrete way but is similar in an abstract, thematic way. The essential point of this theory is not literally to describe how memory is encoded but to build on the principle that every new event that we experience is related to the events of our past experience, and that the events are connected by successive layers of abstraction. Within such a model we can see the emergence of Rosch's basic level categories at the MOP level along with subordinate scenes and superordinate TOPs. There are clear parallels between the conceptual structures Schank describes and the ICMs of Lakoff (1987). The idea of the origin of conceptual models in bodily experience is Lakoff's and I think thi  ( e--itt-t-i-ofound ways of explaining the  Metaphor Comprehension - 11 innate basis for concepts and thinking. Schank does not address such issues. However, the strength of Schank's theory is that he draws upon experiences, rather than concepts, as the source of thought. In Schank's theory experiences become elaborated and compressed into concepts. In other words, Schank's system proceeds from experience to thought to language, although language is not clearly incorporated within it. Lakoff's system arises out of experience, too, but the foundational basis of his system appears to be organized the other way around, with thoughts emerging from language. Where he initially describes concepts like anger as emerging from bodily responses like pressure and heat, he has also discussed the grounding of concepts in social experience - advancing concepts like "arguments are war" or "theories are buildings". In these uses, the word comes to precede and structure the thought. I view this as a serious flaw in Lakoff's arguments. Prototype based models in which language and thought are separate but inextricably bound are finding some support from neurological evidence - Damasio and Damasio (1992) report that neural systems in the left hemisphere represent phonemes, phoneme combinations and syntactic rules, and that other left hemisphere neural systems mediate between these structures and those that categorize nonlanguage representations - such as sensory and motor systems - which mediate between the body and its environment. They have found that specific lesions in these st_ruct_ures can--produ ce -  -  loss of very specific language  Metaphor Comprehension - 12 functions. Damage to the occipital and subcalcarine portions of the left and right lingual gyri cause patients to lose both the ability to perceive and to imagine color. Lesions in other regions can produce other highly specific losses, e.g., the ability to give proper color names, or the ability to say color words. Damasio and Damasio hypothesize that memories about experiences with objects are embodied in ensembles of neurons in 'convergence' regions of the brain. When they are activated, many anatomically separate and widely distributed neuron ensembles fire simultaneously and reconstruct previous patterns of mental activity. Language can be envisioned as a 'sixth sense' - the input that activates the ensemble. The information is categorized so that events and concepts can be reactivated together. These conjectures lend support to prototype and schema models. For example, their ideas could nicely incorporate the kind of generalization and extension suggested by Shank's (1982) MOPS and TOPs, the internalized cognitive models (ICMs) that Lakoff (Lakoff, 1987; Lakoff & Johnson, 1980) proposes, and the PDP models like Chandler's (1991) as well. According to the Damasios, language performs what P. Churchland (in Damasio and Damasio, 1992) has referred to as "cognitive compression" by creating categories that reduce conceptual complexity. They then go on to suggest that because the brain can categorize perceptions and actions simultaneously along many different dimensions, symbolic representations such as metaphor can easily  emerge fr_m_thi-s-arehitectu-r-e.  Metaphor Comprehension - 13 While Schank, Lakoff, and Rosch are all talking about concepts rather than words per se, the line between the two is a difficult one to draw. It seems very likely that when we process language, we are relating what we are trying to understand to concepts derived from incidents of greater or lesser duration and complexity, rather than to verbal definitions, propositions, or feature lists. Hearing the word 'elephant' may only require that we access a scene where we had previously encountered one, while hearing the word 'beautiful' might require access to a TOP extracted from multiple exposures to events where various things have been described or experienced as beautiful. Just as we compress events into stories, so we might compress stories into themes, and themes into words. It is worth noting that some of the neurological evidence offered by Damasio and Damasio (1992) lends strong support to this possibility. Thought, in such a scheme, is the process of extracting episodes from experience, and language is the way of indexing such episodes. Language would then emerge from the same processes as thought, but thought would always underlie, and necessarily precede it. Without some system of language, we would have episodes, but no way to extract their essence, or to label the result of thematic abstraction. One of the most powerful arguments in favour of thought being independent of language is the universal existence of ambiguous and nonliteral language. If language determines thought, then every time we hear a word, the same meaning should occur to_ns--I-f lamrguaq-permits the same word to represent  Metaphor Comprehension - 14 two or more unrelated concepts (bank), related concepts (cold), levels of abstraction (clear), contextually distinguished interpretations (sick), and even diametrically opposite meanings (Great news!), then language must necessarily be representing more than a straightforward correspondence would suggest. In fact, if thought and language were synonymous, we would not need dictionaries at all, since it would be meaningless to talk about one word in terms of others. How would we get the words that seeded the whole procedure? Somewhere we would have to have a thought before a word or the system would be one of infinite regress. A possible solution would be if we accept Lakoff's explanation as to the origin of the meaning of words, but his embodiment idea really only holds for spatial orientation and emotional words that define internal states. It would be hard to imagine how we would metaphorically derive 'lawnmower', 'elf', 'wrench', or 'hamburger' from either physical or social experience without having first encountered an object or experience, and then, having the concept, acquired a word. New words can permit the restructuring of experience and create concepts where none existed before, but not prior to the elements forming the new structure being in place. Of course, every new structure that becomes available then opens itself to include future input. A model like Schank's conceptualizes thought and language as independent, and yet does not demand a symbolic system of internal representation. Instead the representation takes the form ot-themes , gists,_ or stories, based on episodic -  Metaphor Comprehension - 15 events, observed objects, and learned facts that derive ultimately from perceptual processes. The study of nonliteral language usage can uniquely address some of the questions associated with the relationship between language and thought. Nonliteral language is by definition a contextually dependent use of language, and as such immediately challenges theories in which language and thought are identical. It is also incompatible with a theory that requires thought and language to be completely separate systems, since symbols systematically encode semantic content, and metaphor radically alters the content and relationships (especially truth relationships) of logically structured systems. The study of nonliteral language calls attention to the pragmatic distinction Searle (1979) identified between  sentence  meaning and speaker  meaning. It might be said that nonliteral language serves a metametaphorical function in the study of language. By demonstrating that the meanings associated with individual words are much less structured and specific than they superficially appear to be, we come to realize that the same is true of our apparently literal language as well. With such knowledge, a more realistic investigation of this remarkable human ability is possible.  Metaphor Comprehension - 16  THEORIES OF NONLITERAL LANGUAGE COMPREHENSION The earliest philosophers and rhetoricians recognized that figurative language had special powers not characteristic of common speech. To Aristotle, writing in the 4th Century, B.C. (in Ortony, 1979) this meant that figurative language was used to generate an implicit comparison for the purpose of embellishing speech. In Geoffrey of Vinsauf's treatise on rhetorical language, the Poetria Nova (Nims, 1967), written around 1200, he also identifies metaphor as an instrument of hidden comparison, which is, "introduced not under its own aspect but with dissembled mein, as if there were no comparison there at all, but the taking on, one might say, of a new form marvelously engrafted, where the new element fits as securely into the context as if it were born of the theme. . . This type of comparison is more artistic; its use is much more distinguished." (Nims, 1967, p. 25). Both Aristotle and Geoffrey regarded metaphor as a rhetorician's tool that would either embellish or disguise common language to make it more palatable but less reliable. Language used in this way was not to be trusted, as it created false beliefs by equivocation. Implicit in this view is that there are always clear, literal words available to express any meaning, for which metaphor can be substituted if the speaker wants to avoid being direct and (by implication) truthful. ___The__same—be4-ief In Ll le prima dy 61 literal language is -  -  Metaphor Comprehension - 17 evident in most early explorations into the linguistic nature of metaphor, such as Searle's (1979). Searle's intent was to discover how speaker meaning and sentence meaning "come apart" in metaphor. He emphasized that there is not one literal meaning and another metaphorical, indirect, or ironic meaning to a single utterance, but rather only a single meaning, along with the speaker's intention. The literal sentence sets forth a set of truth conditions relative to a particular context. He also denied that a metaphor is just an inadequate or unbalanced feature mapping between what Richards (1936) had labelled the "tenor" and the "vehicle" (the underlying idea, and the word used to convey it, respectively). Instead, Searle derived three steps for the comprehension of any phrase, claiming that hearers must first decide whether a special strategy is needed to interpret an utterance, depending on whether it seems defective in its literal form. They then must search for properties in the topic that are among the values of the metaphorical vehicle, and finally, they must employ strategies that permit the range of the target to be restricted to that intended by the speaker. These goals of Searle merge with those of Clark and Lucy (1975) who examined the dichotomy between speaker meaning and sentence meaning in the processing of indirect requests. Their three stage model [referred to by Glucksberg (1991) as the Standard Pragmatic model] predicts that the primacy of literal language would produce a reaction time disadvantage in the processing of metaphors. Clark and Lucy set forth- their model as follows:  Metaphor Comprehension - 18 First, the listener derives and represents the literal interpretation of the sentence. Second, he then tests this interpretation against the context to see whether it is plausible or not. If it seems appropriate to the context, then it is taken to be the intended meaning, If, however, it does not seem appropriate, either because it contradicts some obvious fact or because it violates a rule of conversation, it is rejected as the intended interpretation. Third, in the case of such a rejection, the literal interpretation is combined with an appropriate rule of conversation, and this leads, by deduction, to the appropriate intended meaning (Clark & Lucy, 1975, p. 58). The implications of this model include both that nonliteral uses should take longer to understand than literal ones and that deriving a satisfactory literal interpretation would preclude ever going beyond it to explore nonliteral implications. Further, the nonliteral interpretation demands a different kind of contextual support. Such clear-cut predictions make the model particularly amenable to empirical investigation. The threestage model depends on the premise maintained by Grice (1975) and Searle (1979) that speech represents a tacit agreement between all parties in a conversation to be both cooperative and rational or else 'appropriateness to the context' could never be unequivocallydeterzdned, Such -a—perspective reinforces the  Metaphor Comprehension - 19 belief that literal language is always truthful while nonliteral language is not. Stage models, while intuitively appealing, have never been strongly supported empirically (Pollio, Fabrizi, Sills, & Smith, 1984). Hoffman and Kemper (1987), in a thorough review of studies that used reaction time to investigate metaphor comprehension, found no clear evidence in support of such a process. Instead, the findings, particularly in the case of well-learned nonliteral uses such as idioms and indirect requests, indicated that these expressions were understood in context at least as fast as their literal paraphrases (e.g., Bobrow & Bell, 1973; Gibbs, 1979; 1980). When idioms and indirect requests are found to take no longer to understand than matched literal expressions, the results are often attributed to lexicalization. Lexicalization is the process by which any word comes to be represented in an individual's vocabulary as a "semantic chunk" (Hoffman & Kemper, 1987) with a standard meaning attached. This process appears to occur with all words we understand and use readily. According to Swinney and Cutler's (1979) Lexical Representation hypothesis, idioms and indirect requests have a stable, communally recognized meaning of their own that takes precedence over the compositional meanings of the individual words that make them up, accounting for faster understanding times than would be the case if meaning were computed by assembling individual word meanings and relationshigs—Lexicalized-idi orns ncome 'greater than the sum -  -  Metaphor Comprehension - 20 of their parts' in the same way that we ignore the sounds of the individual letters in a word like "drought" when we become familiar with it. Where idioms are understood as quickly or even more quickly than their literal paraphrases, it can be attributed to the 'pre-fab' efficiency of the semantic chunk. The intriguing question raised by the Lexical Representation hypothesis is how the lexicon would handle metaphorical multiple meanings associated with single lexical entries. It is likely that very familiar, conventionalized (i.e., dead) metaphors are lexicalized just as idioms are, since their nonliteral meaning has become stabilized in the same way. To access them, context may shape how the lexicon is approached, preselecting among an array of possible meanings for a single word, or the lexicon might yield a singular meaning which is shaped by context afterwards. Swinney (1979) found that immediately after exposure, both the relevant and the contextually inappropriate meanings of ambiguous words were facilitated in a lexical decision task, suggesting that all known meanings of a word are stored together and accessed simultaneously. Tabossi (1988) reported that prior context could determine whether or not the dominant (most frequent) meaning of an ambiguous word would be activated. An inadequately constraining context results in the accessing of both meanings, while a tightly constraining context prevents access to the inappropriate meaning. Recently, Allen, McNeal, and Kvak (1992) reported results suggesting a word frequency advantage-in-lex±cal decrion tasks, indicating that -  Metaphor Comprehension - 21 accessibility within the lexicon might be according to a preestablished order based on previous exposure. Their results show that both phrase familiarity and context can affect how meaning is shaped. Finally, the results of a series of experiments by Jastrzembski (1981) using high and low frequency words presented without context indicated that word frequency and number of meanings independently influenced recognition speed. Words with a large number of alternate meanings and words with higher word frequency counts were recognized faster than words with few alternates or low frequency. Jastrzembski suggested that each meaning of a homograph is represented with a unique logogen. From these diverse studies, it is possible to conclude that either or both word frequency and word ambiguity information could potentially be part of the lexicon, but how the lexicon is ordered and how meanings are accessed within it is clearly far from resolved. Further, it seems unlikely that conventionalization can have occurred for novel uses of words used metaphorically, where new usages arise spontaneously for words that have previously been lexicalized with a different literal meaning. What might the lexicon contribute to helping understand them? If we accept that lexicalization may contribute to the speed of processing nonliteral language, we are still left with the question of why and to what extent the two are related. If nonliteral words and phrases can be stored and accessed just like literal words, what-accounts for their distinction - i.e., what  Metaphor Comprehension - 22 makes a usage unequivocally nonliteral? And if we are using response time as a measure, what can we then say about frequently used, conventionalized (i.e., dead) metaphors or infrequently used (i.e., unfamiliar) literal words? It seems very likely that the process of conventionalization is all that separates a metaphor from a literal word. On the other hand, it could be that literalness and familiarity are simply being confounded, producing results that distort what we might be able to learn about metaphor. The experiments presented in this dissertation represent an attempt to investigate whether the literal/nonliteral distinction can be separated from familiarity differences when response time is used as a measure of understanding. There have been some attempts to explain nonliteral language comprehension that do not involve lexicalization. Glucksberg and Keysar (1990) and Glucksberg (1991) discuss metaphor as a type of class-inclusion statement. In this interpretation, an "a is b" metaphor produces the attribution of a subset of the features of "b" to "a", so that "a" is consequently included in a hierarchically superordinate category of which "b" is a prototypical exemplar. In this approach, the fact that metaphors are implicit class-inclusion statements is the cue by which we recognize that a usage is metaphorical. This hypothesis provides a plausible model for how metaphors are comprehended. In particular, it suggests how aspects of the literal meaning can come to be selected_^-14^  e nonliteral meaning of the  Metaphor Comprehension - 23 same word. Further, it accounts for the familiar observation that metaphor is often employed when a situation is new, or when existing words seem inadequate. Perhaps the strongest advantage of Glucksberg and Keysar's (1990; Glucksberg, 1991) model is that it can be developed to provide an explanation for the distinction between dead metaphors and novel ones. Identifying and constraining a new superordinate category could take time, but repeated exposure to the same usage would lead to the superordinate category, named by a prototypical member, being available with the same speed as the exemplar itself. The category name would in effect shift to the broader superordinate category, encompassing both. For example, if the primary meaning of 'clear' had been 'visually transparent', use of the word to indicate 'completely understandable' could require an ad hoc cognitive category construction in its initial use. Following repeated usages, the hierarchical category would become more flexible, and eventually both meanings could be subsumed under the same word. It is not really that there are two distinct meanings accessed in parallel, but rather a broader single meaning that context can shape to its purpose. Another way to explain metaphor comprehension is to place the responsibility for shaping meaning largely in the expectations generated by context. This is the approach suggested by Ortony, Schallert, Reynolds, and Antos (1978) who conducted two experiments to investigate context effects and  Imsxsing time differences fbr idioms and metaphors as opposed  Metaphor Comprehension - 24 to words used literally. In their first experiment, subjects were given either long and short versions of contexts followed by a target phrase which had a literal interpretation in one context and a metaphorical interpretation in another. As soon as they had read and felt they understood the target, they were to press a response bar. Ortony et al. found that in these conditions, metaphors and their literal paraphrases showed no statistically significant differences in time to understand target sentences when preceded by long (mean = 45 words) contexts. When contexts were short (mean = 6 words), interpretation time was slower for all conditions, but metaphors were slowed to a significantly greater extent than literal words. In their second experiment, idioms rather than metaphors followed intermediate length contexts (mean = 24 words) under three conditions. Two conditions presented the context with a target ending consisting of either the idiom or a literal paraphrase of the idiom, and the third condition used a different context in which the idiom was given a literal interpretation. The results of this experiment indicated that idioms took no longer to process than either their paraphrases or the same words used literally. From these data, they concluded that context was the key to understanding metaphors. As linguistic material is processed, various schemata (Rumelhart and Ortony, 1977) produce expectations that are fulfilled by both literal and metaphorical targets. The main difference is that there is more consistency in evoked schemata between context and^ target in the case of literal endings than  Metaphor Comprehension - 25 metaphorical ones. The degree of consistency between the evoked schemata of the context and the target accounts for the speed (and, presumably, reliability) of understanding. By emphasizing the importance of context, Ortony et al. (1978) have drawn attention to a factor that must play a vital role in constraining the meaning of words used metaphorically. Acknowledging their claim that given inadequate context it is often difficult to understand metaphors and given substantial context they are understood much more readily, the same could be said about a visually or auditorially degraded word, or a completely absent one. We can often understand metaphors given a very underdeveloped context ("Some marriages are iceboxes") and at other times the consistent schematas between context and target may be very tenuous ("The dawn came up like thunder") and yet the metaphor has considerable cognitive power. Still other metaphors may generate very consistent expectations and yet convey little meaning ("My surgeon is a nurse"). It is also important to realize that often the clarifying context can come after a metaphor or from extra-linguistic sources. It does not have to precede the metaphorical statement to make it interpretable. Finally and most importantly, the function of metaphor in communication is not illuminated by attributing its interpretation solely to schemata generated by the context. Why would we ever choose a word to complete a context that "leaves over" some incompatible aspects, when we apparently have some choices AW5ri4 iiferal words that are more fully -  -  Metaphor Comprehension - 26 consistent? And why would we ever tack on an obscure statement to reinforce a substantially established one? Any adequate explanation of metaphor must address these enigmas. Finally, some investigatory metaphor research has been attempted through the use of connectionist, or Parallel Distributed Processes (PDP) models (e.g., Chandler, 1991; Hubbell, 1988). Chandler (1991) argued that there is no difference at all in the processing of metaphors and literal words, though he acknowledges a psychological reality to the distinction. He cited evidence that metaphors presented in context show no more processing time than literal controls. Like Ortony et al. (1978), Chandler emphasized that context must make an important contribution to shaping how any word is understood. What Chandler attempted to convey is that all words are far more 'open' than we realize. Even when used literally, our understanding of what is being conveyed is shaped by sentential and extrasentential context to such an extent that we can hear "Get your dog out of the middle of the floor" and let "dog" refer to a brown, animate, 30 kg. mammal in one case and a pink, inanimate, 300 g. toy in another. Virtually nothing except a vague similarity of shape may unite the two. Metaphorical uses of words, he argues, can be accommodated similarly. Chandler conjectures that recurring patterns of input features tend to become organized into neural patterns of associations, with some characteristics having stronger associations due to more constant activation. Lm-a- simplified summary, metaphors will result in  Metaphor Comprehension - 27 the strong activation of both the expectations of context and the representation of the word used metaphorically in which incompatible features have been inhibited. No distinctly literal interpretation is ever made, but literal and metaphorical are determined by whether a substantial proportion of microfeatures in the vehicle need to be inhibited in order to coincide with the developing context. In this dissertation I attempt to build on the strengths of Ortony et al.'s (1978) Glucksberg and Keysar's (1990) and Chandler's (1991) models of metaphor understanding. All three independently predict that metaphors as a class can be understood just as quickly as literal words as a class. However, making a distinction based on the conventionalization of a metaphorical use may affect this prediction. Novel metaphors should take longer to understand than dead metaphors or literal words, because it should take more time to constrain and construct a new meaning than to simply apply a category label. The parallel case to novel metaphors would be unfamiliar literal words, where meaning would also need to be constructed from a combination of what might be known and/or available in the word itself, and what can be gleaned from context. This effort should likewise be revealed by longer response times. In natural situations, context is a large part of the constraint process, and a process like the one Glucksberg and Keysar (1990; Glucksberg, 1991) describe could not realistically CP/Astra-in, a_metapher untess uidAt4ief were applied. Although I -  Metaphor Comprehension - 28 will argue that while context probably cannot completely determine meaning, deriving the full intended meaning of a metaphor is likewise impossible without considering it. The ability to develop a richer meaning through the combination of contextual constraint and familiar meaning provides an efficient and enlightening access to understanding. This dissertation explores more precisely the contribution of context to understanding, both by manipulating the expectations generated by context and by manipulating the familiarity of words in the same context. Chandler's (1991) connectionist interpretation incorporates contextual effects and could support a model such as Glucksberg and Keysar (1990; Glucksberg, 1991) propose, but it is reasonable to conjecture that a connectionist system should respond more quickly to activations of established connections than to exceptional ones, where a period of interactive competition among microfeatures is required for interpreting anomalous input. That is, a PDP system should take more time to understand novel metaphors. Both approaches (Glucksberg, 1991; Glucksberg and Keysar, 1990; and Chandler, 1991) acknowledge a psychological reality to the nonliteral/literal distinction that draws on the unaddressed features of the metaphorical vehicle. Metaphors are almost universally regarded as adding an ineffable depth that is not conveyed by a literal paraphrase. This seems an essential point in regard to understanding why metaphor exists, and how it functions. Even the rhetorical implications o_f_metaphor - however) appear to lie not in metaphoricity but in  Metaphor Comprehension - 29 conventionalization. A long-dead metaphor probably has little more rhetorical force than its literal paraphrase (which, in many cases, is only a longer-dead metaphor) but a newly introduced metaphor resonates with the connotations of its source and use. Given the state of understanding about metaphor comprehension in the psychological literature, a number of questions remain unanswered. None of the major models of metaphor understanding is without drawbacks. It is my aim in this dissertation to develop a viable model of metaphor understanding that explains and unifies previous results. To do this I investigate several questions that have shaped the study of metaphor over the last 15 years, including whether or not metaphor understanding requires a different cognitive process than literal word understanding and whether response time differences can be considered a valid indicator of such a process. A major emphasis is to explore the effects of written word frequency (as a measure of conventionalization) on response times for metaphor understanding, and to show how the differences in 'time to understand' for conventionalized as opposed to unconventionalized words are often, and mistakenly, taken as evidence of a literal/metaphorical distinction. The most difficult goal of the dissertation, however, is to begin to provide an answer to the question of why metaphors exist at all. Specifically, the experiments reported here investigate the following hypotheses: al That-t-he-aeeepLed digtinction between literal and  Metaphor Comprehension - 30 nonliteral as separate kinds of words is not a viable one. Words that we designate as literal may have origins in other uses that have become modified over time. Literal meaning is generally associated with culturally dominant meaning, and more likely reflects how frequently users are exposed to a particular usage of the word. (2) That even if the dichotomy commonly recognized between literal and nonliteral is untenable, there is a difference of degree between using a word in a conventionalized way and extending the meaning of the word in a way that makes use of some of its prototypical characteristics to expand a conceptual class. The difference is that something new is being conveyed, and receiving this new information takes time. The compensation is that through this process, metaphor derives the denotative and connotative force that contributes to its rhetorical impact. Unfamiliar literal words that undergo the same process will take longer and suffer in comprehension because the bootstrapping structure is either considerably weaker or missing. (3) That context is an essential component in constraining novel metaphors. Context does not consist solely of the words immediately preceding the metaphor. It may develop from some previously established understanding between speakers, some nonlinguistic sensory information present at the time of the exchange, or some words used during the exchange. Without context, words are likely interpreted according to their most  Metaphor Comprehension - 31 (4) That the lexicalized meaning of a word used metaphorically is an essential component in deriving meaning from context. Without considering the meaning carried by the literal usage of the metaphor, context has only a stochastic function, allowing us to formulate a prediction of what is to come. The information provided by the following words is already anticipated, and the information conveyed is limited. Using words metaphorically goes beyond the expectations generated by context to expand and sharpen communication, often in ways that are thought provoking and enlightening. Although there is some disagreement about the extent to which context is required for the determination of literal meaning (see Dascal 1987, 1989; Gibbs, 1984, 1989; Searle, 1979), it seems essential that both the context and the word used metaphorically must contribute jointly to understanding nonliteral language. Simply providing a highly developed context is not enough, nor is the metaphorical word in isolation. In these experiments I attempt to demonstrate how the two elements work together by comparing conditions where contextual expectations are supported with those where they are violated.  Metaphor Comprehension - 32  INTRODUCTION TO THE EXPERIMENTS Six experiments are presented. The first compares idioms and metaphors to their literal paraphrases to pursue the findings of Ortony et al. (1978) that nonliteral metaphors and idioms do not necessarily take longer to understand than their literal paraphrases. The presentation is divided into blocks of alternately presented idiom and metaphor portions, and for each, the three conditions used in Ortony et al.'s idiom experiment are adapted for use here. The example of an idiom item provided in their article (p. 471) is as follows:  Idiom version Context: Dean spoiled the surprise that Joan had been planning for their mother's birthday party. When he realized what he'd done, he apologized for having Target: let the cat out of the bag. Literal version Context: Walking back from the store, Anne found a kitten which she put in with her groceries. She got home and her puppy went wild when she Target: let the cat out of the bag. Control version Context: Dean spoiled the surprise that Joan had been planning for their mother's birthday party. When he realized what he'd done, he apologized for having Target: revealed the secret. As the other stimulus items used in their original studies were not available, an entirely new set of stimuli was developed for this experiment. Thus, an experiment based on their design  w_as_ nPc- sary it order to establish the efficacy of the new -  Metaphor Comprehension - 33 stimuli. The idiom portion of this experiment was derived from the idiom portion of Ortony et al., and the metaphor portion used the same presentation technique with single word metaphors. All contexts in both portions were similar in length to those of Ortony et al.'s idiom experiment. Subjects were presented with items from both portions in alternating sets of 20 items per set. The same context was used for each stimulus item in every condition except for the LITERAL condition, where the same target word(s) presented in the NONLITERAL condition were used literally. These naturally required different contexts in order to evoke the literal meanings of the same word(s). The three conditions adapted from Ortony et al. (1978) and used for both portions of the experiment were: NONLITERAL, where the target was either an idiomatic phrase (replaced by a single word dead metaphor in later studies) or a single word metaphor (replaced by a single word novel metaphor in later studies); PARAPHRASE, where a literal word target with the same meaning as the NONLITERAL target was presented in the same context; and LITERAL, where the same word as the NONLITERAL target was presented in a different context. Sample items are provided with the experiments to follow, and the Appendices contain a complete listing of the stimuli and targets from all six experiments. In addition to the NONLITERAL, LITERAL, and PARAPHRASE ending conditions employed by Ortony et al. (1978), the present series of experiments also included a SURPRISE condition in which literal word-en-ellilgs that Were plausible and grammatical but that  Metaphor Comprehension - 34 violated contextual expectations were included. The SURPRISE endings were primarily used to investigate whether context plus literal word understanding is inevitably faster than context plus metaphor, or if the interaction of contextual expectation with target word meaning also plays an important part in how readily understanding occurs. It was felt that there should at least be a measurable difference between the PARAPHRASE and SURPRISE results if the stimuli were valid. As the results will show, the stimuli were extremely sensitive to differences between conditions, but the SURPRISE condition proved to be even more valuable for the additional perspective it provided about the relative effects of context and literal words. The second experiment replicates the first with dead metaphors and novel metaphors rather than idioms and metaphors. Its importance lies in its ability to demonstrate that it is unrealistic to generalize conclusions about metaphorical processing unless the relative conventionalization of the metaphors is also considered. In order to postulate a special process for understanding metaphors, such a process must work on all metaphors in the same way. If dead metaphors can be shown to take less time to understand than novel ones do, it is untenable to simply conclude that there is a special, time-sensitive cognitive process that interprets metaphors. The third experiment replicated the second one, with two important refinements. First, an UNFAMILIAR condition was added to both pnrtians- f - which umtAhiliar literal words with the same -  Metaphor Comprehension - 35 meaning as the more familiar NONLITERAL and PARAPHRASE targets were presented in the same contexts. Second, an attempt was made to equate the overall printed word frequency for all the conditions according to Francis and Kucera's Frequency Analyses of English Usage: Lexicon and Grammar (1982) - except, of course, for the UNFAMILIAR endings, where printed word frequency was kept as low as possible. This experiment was designed to show, first, that unfamiliar words used literally could take at least as long to process as familiar words used nonliterally, and that response time results could be explained as a function of the conventionalization of the word with a particular meaning attached, independent of whether the word conveying that meaning was used literally or nonliterally. The fourth experiment represented a departure from the previous line. Using the same stimulus items as in Experiment 3, subjects were presented with a sample of 20 of the sentence contexts and asked to generate their own single word endings. This experiment was motivated by the results of the earlier experiments, which showed that some metaphors could be understood as quickly as literal words. It asked whether subjects could then also be expected to generate metaphors as freely as literal words or whether words generated had any relationship to the subject's familiarity with their meanings. It also looked at intersubject consistency with respect to word generation. Experiment 5 returned to the read-and-respond presentation tonvertry thb earlier results in a more -  -  Metaphor Comprehension - 36 discriminating response paradigm. Rather than pushing a button when the meaning of the target was ostensibly 'understood', subjects were presented with two words that had been matched for length, word frequency and grammatical appropriateness and asked to choose the one that made more sense in terms of the context. In addition to latency, this response paradigm provided data regarding errors, giving a clearer indication of when targets are misunderstood. Errors can be as useful as latency information with regard to how context and target function together. The final experiment, Experiment 6, reduced context to a minimal level in order to assess how much understanding can be attributed to specific elements of information contained within the context, and how much is contained in, or is recoverable from, the target word. As it was matched to Experiment 5 with regard to items and presentation, a further series of analyses provided comparative data for Experiments 5 and 6 taken together. A fundamental theoretical assumption concerning the nature of literal and nonliteral language was made for the purposes of constructing the stimuli for this dissertation. Thus far, the words 'literal' and 'nonliteral' have been used as is typical in research on this topic in psychology, which is to say, without establishing any clear definition of them. However, one intention of these experiments is specifically to investigate whether or not any differences exist at all between literal and nonliteral and, if so, where the line that separates them should be drawn. In-a---previous study in which reaction time differences -  Metaphor Comprehension - 37 based on familiarity were found, Blank (1988) went so far as to identify four classes of metaphors - very familiar, fairly familiar, transitional, and novel - without specifying any objective criterion (beyond their source) for defining their limits. In order to impose a working definition by which to distinguish literal from nonliteral for the experiments in this dissertation, 'literal' is used to denote the most conventional usage of a word in a cultural sense (i.e., the definition accorded primary status in an established dictionary). All secondary usages are taken to be nonliteral. The distinction drawn by Searle (1979) between word meaning and speaker's utterance meaning emphasizes the fact that there can never be a 'stand-alone' metaphor; i.e., one that does not make use of a word already in the lexicon. The same word can carry a variety of meanings, but the meanings associated with the words vary in the frequency with which they are employed. All of the meanings associated with a word have some elements in common, but it is possible that only the most conventional usage of the word can be called literal. True homonyms, with completely different etymological origins and independent meanings, such as 'bank' as a savings institution and 'bank' as a river's edge, represent a different case. 'Bank' in this case does not have one literal and one metaphorical meaning. It has two literal meanings, and each can independently be used as a metaphor. The criterion of conventionality has likewise been given an definitturF fo  ---  _ ------th e purposes of the experiments in this -  Metaphor Comprehension - 38 dissertation by equating it with printed word frequency. This choice is based on the assumption that literal meanings can be viewed as the ones that are most often conventionalized among members of a linguistic community, although it is recognized that there will always be differences in both the number and order of lexicalized meanings among individuals. It is readily acknowledged that the cognitive lexicon may have both a form and a representation of language that does not reflect frequency of exposure, or that encodes such information much differently. Word frequency counts were only employed in these experiments as a way to control for the variability that such exposure might produce. A cautionary note that underlies all of the research upon which this thesis is based is also in order here. The site of these experiments, Vancouver, British Columbia, Canada, is rapidly becoming a fully multicultural community, and the University from which research subjects are drawn reflects that cultural diversity. Many students are still in the process of attaining full fluency in English. Students who learn English as a second language often learn it in a different fashion than native speakers of a language, and one of the major differences is that second language learners usually learn syntax and vocabulary independently. Vocabulary is dictionary based, and thus the language is learned in a bottom-up fashion. Conventionalized expressions such as idioms have to be acquired separately. Not—surprisingly, these phrases often seem unusual  Metaphor Comprehension - 39 and can present a great deal of difficulty to new language learners. Such evidence strongly favors one of the major premises of this dissertation, that idioms and indirect speech acts are learned through exposure and then lexicalized, while novel metaphors are not. The experiments presented here represent an attempt to reach a compromise between an awareness of the limitations non-native language speakers have, and a desire to reflect a representative sampling of the student population. In particular, since students were participating in the experiments for class credit, it seemed unjust to discriminate against some potential subjects on the basis of so elusive a quality as English language fluency. However, it soon became apparent that fluency in English was an essential component to validating the experiments. The solution was to note on the signup sheet that fluency in English was necessary, and allow students to self-select on that basis. Afterwards, a performance criterion was imposed, both in the data (latency) and post-test (comprehension) portions, to screen out subjects (regardless of native language) whose scores were two significantly beyond the experimental group norm. This took the form of first excluding from analysis any response time under 100 msec or over 5,000 msec in Experiments 1, 2, and 3, or over 10,000 msec in Experiments 5 and 6. In addition, any individual response time more than two standard deviations beyond any  .  subject's condition mean was not included in that mean, and any subject whose condition—means exceeded The group condition means  Metaphor Comprehension - 40 by two standard deviations in more than half of the conditions was dropped from the analysis and replaced. These measures served to protect the data from persons who claimed English fluency but had less familiarity with the language than other subjects to contribute data. There is no doubt that cleaner results could have been attained if the experiment had been limited to only third generation English language users, but since such a group no longer represents the social reality of this community, the experiments presented here should be understood to represent results derived from competent readers with a wide range of linguistic backgrounds.  Metaphor Comprehension - 41  EXPERIMENT 1 - PROCESSING IDIOMS AND METAPHORS Numerous empirical studies (e.g. Bowbrow and Bell, 1973; Clark and Lucy, 1975; Gibbs, 1979; Ortony, 1978; Swinney and Cutler, 1979) have supported the contention that some forms of nonliteral language, such as idioms and polite expressions (indirect speech acts), are processed in context at least as fast as their literal paraphrases. Experiment 1 adapted and extended the findings of one such experiment (Ortony, et al., 1978) with regard to time for processing different kinds of target endings, and also attempted to isolate more precisely how much subjects rely on context to shape understanding. The stimuli consisted of a series of items having .a similar form and presentation to those in Ortony et al.'s (1978) idiom experiment. In their paper, the results of two experiments were discussed. The time required to understand metaphors was measured in the first experiment where context length was manipulated, while in the second, time to understand idioms was measured and context remained constant at an intermediate length. Metaphors, as opposed to idioms, were not investigated in intermediate length contexts. One aim of the present experiment was to investigate whether single word metaphors would take no longer than their literal paraphrases to understand when context length was controlled, or if Ortony et al.'s finding of no statistically significant difference applied only in the case of well ie  CiL  1  ed, conventionalized idioms, or  Metaphor Comprehension - 42 metaphors in well developed contexts. A second goal was to investigate Ortony et al.'s claim that context length per se is a factor in how quickly metaphors are understood. The functioning of context was addressed by adding a fourth SURPRISE condition to the three of Ortony et al.'s (1978) idiom experiment. The SURPRISE condition resembled the situation described by Garrod and Sanford (1988) about the consequences for text interpretation when role casting assumptions are violated. SURPRISE endings were required to be grammatically correct and plausible, but not as likely to be anticipated from the context. Three reasons motivated the inclusion of the SURPRISE condition in these experiments. First, its endings were all literal uses of words displayed in the same context as the PARAPHRASE endings. Thus a direct comparison could be made between response times for an expected and an unexpected literal word, to investigate how much of a contribution contextual expectations could make to understanding when nonliteral usages were not involved at all. While it is naive to expect that every word used literally would take the same amount of time to process, it is common among researchers when dealing with literal and nonliteral uses of words to treat them as if each example represented an undifferentiated class. By applying this assumption of literal and nonliteral word homogeneity, processing time differences between the conditions can be ascribed to the differences produced by meeting or thwarting contextual __exp ecta tions... Second, reponse time f6r SURPRISE endings can be -  Metaphor Comprehension - 43 compared to NONLITERAL endings for both metaphors and idioms, to see the amount of disturbance processing nonliteral words would  cause relative to unexpected literal words. The third reason was a purely methodological one, since the predicted result of no difference between literal and nonliteral targets would only confirm the null hypothesis and leave the experiment open to the contention that perhaps the stimuli were simply ineffectual. The SURPRISE condition was expected to slow response times relative to all other conditions. Each stimulus item consisted of four conditions. These were created separately for idioms and metaphors. The conditions were: NONLITERAL - where the target was either an idiom or a metaphor, PARAPHRASE - where the target was a literal word whose meaning was similar to that of the nonliteral target word, LITERAL - where the nonliteral word or phrase appeared in a different context that required it to be understood literally, and SURPRISE - where the target was a literal word that was not likely to be anticipated (i.e., was surprising) given the context. Targets within a stimulus item were matched as closely as possible for length. Contexts for each stimulus in all but the LITERAL condition were identical. The dependent variable throughout the experiment was the response time to understand the target ending, given unlimited self-paced exposure to the previous context. An example of an idiom item from Experiment 1 is shown in Figure 1. iT^the-founcandItions  Within each portion allowed a  Metaphor Comprehension - 44  CONTEXT  Chris and his friends were fooling around when they accidentally broke his mother's treasured crystal vase. Several days passed before his mother noticed, and by then Chris was relieved to  ENDINGS  [NONLITERAL] get it off his chest. [PARAPHRASE] confess his guilt. [SURPRISE] get his cuts treated.  LITERAL CONTEXT  Elmer was working on his car when the jack slipped, pinning him underneath. He was not seriously injured, but it took three rescue workers to  ENDING  [LITERAL] get it off his chest.  p e iom item from Experiment 1.  Metaphor Comprehension - 45 number of useful comparisons of response times. Comparing NONLITERAL to PARAPHRASE provided insight into how long it takes to understand the same meaning when expressed nonliterally and literally. Comparing NONLITERAL to LITERAL gave a measure of how long it takes to understand the same word or phrase used nonliterally and literally. Comparing PARAPHRASE to SURPRISE provides a contrast of how long it takes to understand a literal word or phrase with a likely meaning and a literal word or phrase with an unlikely meaning in the same context. And the NONLITERAL to SURPRISE comparison examines the relative disruption created by two possibly different cognitive effects. Method Subjects Thirty-two undergraduates were recruited from the University of British Columbia Psychology Department subject pool. Each received one class credit for an hour of participation. All indicated that they were competent readers and speakers of English. Two subjects were dropped and subsequently replaced when a preliminary tabulation indicated that their overall response rates were more than two standard deviations slower than the combined means for all subjects in over half of the conditions. Stimuli and Procedure Subjects were given an introduction sheet to explain the experiment before it began. They were seated in front of a _Ifew letf___Packard Neetra ES 7 12 computer with a standard color -  -  -  -  Metaphor Comprehension - 46 monitor in monochrome mode and introduced to the response procedure. A mechanical push button connected to the computer timed the responses to within one millisecond (msec) accuracy. Five unrecorded practice trials preceded the data collection. Subjects were told that a post-test consisting of a written, true/false recall questionnaire would follow the visual presentation so it was important to actually read and understand the items, not just to respond quickly. To ensure that contextual effects were consistent for each presentation, it was stressed to subjects that the success of the experiment depended on their reading a context each time it recurred, even if it was recognized from a previous trial. Idiom and metaphor items were constructed in the same way, with similar length contexts of one or two sentences for both types of item. Contexts varied in length from 22 to 39 words in the case of idioms and 18 to 27 words for metaphors. The idiom targets were 3 to 7 words in length, while metaphors were generally a single word. All stimulus contexts and targets, and a copy of the post-test, are provided in Appendix A. Subjects were presented with an entire context at once, which remained on the screen until they had read and understood it. They then pressed a button to display the target. When they had read the target and understood its meaning in terms of the sentence context that preceded it, they pressed the same button as quickly as possible. Twenty different idiom items and twenty different metaphor it_ens_were-presented7  -  ach hAVIE4 four different -  Metaphor Comprehension - 47 targets, for a total of 160 trials per subject. Trials were grouped into eight sets of twenty stimuli with either idiom or metaphor items as targets (4 sets of each). The order of sets was counterbalanced, with each subject receiving a different order, but idiom and metaphor sets always alternated. Each set had one condition (e.g., NONLITERAL, PARAPHRASE, LITERAL, SURPRISE) of each item, and an equal number of the same condition (5 of each). The order of presentation of items was randomized within each set. After all the responses were made, subjects were given a post-test consisting of a paper-and-pencil questionnaire with 32 true/false items to be recalled from the 160 presentations. The items were designed to test whether a specific meaning of the completed presentation was retained correctly. Corrections for violations of the statistical sphericity assumption for repeated measures analyses of variance (ANOVA) were calculated for all data reported here and in the experiments that follow. Whenever degrees of freedom or significance levels are reported in any of the following experiments, they reflect values corrected by applying a Huynh-Feldt epsilon factor to the experimentally derived degrees of freedom (Kirk, 1982). The SURPRISE condition represented a somewhat different situation than the three other conditions, which shared either context or target with one another and so directly addressed the literal/nonliteral question. Its longer response times were enotg lh to pr^o- • - icily violations, and it was felt  Metaphor Comprehension - 48 that a more meaningful interaction analysis would involve just the NONLITERAL, LITERAL, and PARAPHRASE conditions. These are referred to as the base conditions. Two primary analyses - one for all conditions and one for only the three base conditions are reported in all the following experiments, but subsequent analyses are for base conditions only. Throughout, response latencies for all the conditions are supplied in the tables and illustrated in the figures. The conditions in these experiments constitute a withinsubjects factor, so that the same context is repeated three times for each subject (four times in Experiments 3, 5, and 6) with a different target each time. The decision to repeat contexts was made because it was felt that inter-subject variability in the understanding of nonliteral language was likely to be very high in experiments of this type, and a repeated measures analysis with a sphericity correction would be a better way to control error variance than a between-subjects design. The tradeoff in this choice is that with repeated presentations of context, priming effects can occur. Since order of presentation of context sets was different for every subject, it was felt that any priming effects would be distributed evenly over conditions. However, as an additional check against the possibility of such effects, analyses by order of presentation were performed. This was done by combining all the first, second, third and fourth (where applicable) exposures to a context type, and calculating a mean score for each -then applying the same criterion of  Metaphor Comprehension - 49 deviation as for condition means, by eliminating all scores under 100 msec, over 5,000 msec (10,000 msec in Experiments 5 and 6), or beyond two standard deviations. ANOVA was then calculated to determine whether there were any significant differences between first exposures and last exposures to the same context in any condition. Order ANOVAs are reported only for the three base conditions. Although the consistent finding of no significant differences in order effects in this and subsequent analyses can be challenged as indicative of lack of power in the test, visual inspection of order means lends support to the premise that priming as the result of order of presentation have not created spurious effects in the data. First presentations appear to show slightly longer response times in most of the conditions analyzed, but this effect is generally consistent over the various conditions and the pattern of the data is consistent with that of the overall means for both the first and last presentations. In any event, presentation order could not have produced condition differences, since order of presentation was randomized over conditions and between experiments. Further, the consistently faster responses shown by the LITERAL condition (where contexts appeared only once) when compared to the PARAPHRASE condition in all but Experiment 6 make it unlikely that priming can be cited as an explanation for the effects obtained in these experiments. Results throughout report only order_by portion  by -cond±tturr interactions.  Metaphor Comprehension - 50 Finally, the data were also analyzed by dividing the stimulus contexts in each experiment according to their arbitrarily assigned numbers into groups of even and odd, and performing an analysis of variance with these groups as a factor to provide some assessment of stimulus generality. Outliers were dropped from these groupings in the same manner as in the order analysis. While it must be stressed that the nature of stimuli of this type is inherently singular, it is true that the stimuli should be at least generally representative, and this seemed a reasonable check to impose. ANOVAs by half (even/odd) are reported only for the half by portion by condition interaction, and only for the three base conditions. Results and Discussion Results on the post-test for all subjects ranged from 81100% correct, with a mean of 91.5%, indicating that subjects had read the contexts of all conditions to the point of understanding. Responses for any individual target that differed by more than two standard deviations from that individual's mean for the condition were dropped and the mean recalculated based on the remaining responses. Responses of less than 100 msec or more than 5,000 msec were also eliminated. These discarded responses included obvious keypress errors at the low end, and cases where a target was either incomprehensible or unknown to the subject (as in the case of idioms) at the high end, and amounted to less than 5% of the _total data.  -  Neap response times in msec for every -  Metaphor Comprehension - 51 condition of this and every subsequent experiment are provided in Table 1, along with their standard errors. A 2 X 4 repeated measures analysis of variance was conducted on the mean response times for all items in the four conditions for both portions (idioms and metaphors). The two-portion, fourcondition overall ANOVA revealed a statistically significant effect for portion [F(1,31) = 171.45, p < .001] and condition [F(1.4,42.6) = 46.55, p < .001], and a significant portion by condition interaction [F(2.6,80.3) = 17.47, p < .001]. The statistical significance of the portion result is attributable to the reading time differences, given the fact that the metaphors were mostly single words, while the idioms were several words long. Mean response times obtained in each condition appear in Figure 2. In the three-condition analysis, the effects of portion and condition are significant at the same level as for the fourcondition analysis, but the interaction of portion and condition is no longer statistically significant when the SURPRISE condition is excluded [F(1.9,58.4) = 1.35, p = .267]. These results do not support the conclusions of Ortony et al. (1978) and others, that idioms and metaphors presented in context take no longer to understand than their literal paraphrases or than the same words used nonliterally. As Figure 2 shows, responses in the NONLITERAL condition for both portions are consistently slower than for the PARAPHRASE and LITERAL conditions. Possible_pr-ia-i-ng-ef fects dUd tO repeated presentations of -  —  -  Metaphor Comprehension -^52 TABLE 1. CONDITION MEANS (IN MSEC) FOR FIVE EXPERIMENTS PORTION  CONDITION  EXP 1  EXP 2  EXP 3  DEAD  NONLITERAL  1092*  765  844  1115**  1144**  DEAD  PARAPHRASE  1029*  780  849  1197**  1178**  DEAD  LITERAL  1041*  730  766  1024**  1138**  DEAD  SURPRISE  1341*  998  963  1675**  1475**  DEAD  UNFAMILIAR  1165  1592**  1648**  NOVEL  NONLITERAL  783  936  922  1503**  1589**  NOVEL  PARAPHRASE  719  805  828  1136**  1182**  NOVEL  LITERAL  688  737  818  1049**  1349**  NOVEL  SURPRISE  840  940  1017  1848**  1546**  NOVEL  UNFAMILIAR  1214  1767**  1789**  EXP 5  EXP 6  STANDARD ERRORS (IN MSEC) FOR FIVE EXPERIMENTS DEAD  NONLITERAL  51.8*^38.0  59.9  39.1**  31.5**  DEAD  PARAPHRASE  43.5*  35.4  56.8  55.7**  42.0**  DEAD  LITERAL  47.4*  34.1  52.5  39.8**  51.9**  DEAD  SURPRISE  72.7*  61.5  78.1  99.0**  72.9**  DEAD  UNFAMILIAR  94.2  94.6**  70.4**  NOVEL  NONLITERAL  38.4  59.0  71.5  112.3**  71.1**  NOVEL  PARAPHRASE  31.3  37.7  52.2  38.5**  36.9**  NOVEL  LITERAL  29.7  35.2  58.8  35.3**  55.9**  NOVEL  SURPRISE  44.5  56.9  80.1  147.4**  79.2**  NOVEL  UNFAMILIAR  105.9  117.4**  74.7**  * Idioms rather than dead metaphors. ** Semantic _decis-i-on- rather than time to understand. -  Metaphor Comprehension - 53 the same stimuli were investigated in the order analysis. This analysis looked at whether response times for first-presented contexts differed significantly from times for last presented contexts. The ANOVA of the three base conditions found no statistically significant order by portion by condition interaction [F(1.7,51.4) = .94, p = .381]. Generalizability of contexts was investigated by comparing response times for even numbered contexts to those for oddnumbered ones. In this analysis, a significant group by portion by condition interaction was found [F(1.4,44.3) = 5.16, p = .018]. Odd numbered items were consistently faster than even numbered items for both portions, even though the portions were completely independent. While no explanation can be made to account for this result, it does sound a warning about the vagaries of linguistic stimuli, and the need to obtain strong, consistent effects before any conclusions can be drawn. Generalizing over the results of both of their experiments, Ortony et al. (1978) concluded that "What determines the difficulty of processing is not literalness but relatedness to context" (p. 475). The present results qualify such a contention. While literalness does not produce a unilateral effect, since response times to the same context were both shorter (in the PARAPHRASE condition) and longer (in the SURPRISE condition) than NONLITERAL responses, it does appear to have a reliable effect. The base condition analysis shows a statistically significant effect for condition [F(1.9,58.4) =  Metaphor Comprehension - 54  MEANS - EXPERIMENT 1  1400  Z3 (.1)  1200  2 ILI  z  1000  0 tx 800 METAPHOR I IDIOM  600 CONDITIONS - EXPERIMENT 1  Figure 2: Results of Experiment 1 - Idiom and metaphor portions. (N) nonliteral fidio_mimetaphori (P) paraphrase (U literal (S) surprise.  Metaphor Comprehension - 55 24.32, p <.001]. While responses take longer when the context/target relatedness is disrupted in the SURPRISE conditions, just as Ortony et al. contended, slower response times also occur for the NONLITERAL conditions. An examination of the data of Ortony et al.'s (1978) metaphor experiment suggests that their data show a similar pattern to those obtained in the present experiment, despite marginal statistical results indicating no significant difference in their long context condition [F(1,66) = 2.73, p > .10].^A rough estimation derived from their Figure 1 graph (p. 470) indicates that even when metaphorical phrases were presented following a long context they took approximately 250 msec longer than their comparable paraphrases. Their long context condition produced an estimated ratio of difference/overall mean response time results of 250/1950 msec, or 12.8%, while the same comparison for the metaphor portion of the present experiment produced a ratio of 95/735 msec, or 12.9%. It is possible that the only reason the differences were not statistically reliable in Ortony et al.'s experiment was that reading an entire metaphorical phrase produced more variability than reading a single word metaphor. In addition, variability in their results was higher because analyses were based on between-group samples with only 16 trials per group. The findings of the idiom portion of this experiment are also at odds with Ortony et al.'s (1978) conclusions. They found no _significant difference between idioms and controls (NONLITERAL -  -  Metaphor Comprehension - 56 and PARAPHRASE conditions here), while LITERAL uses of idioms took longer. In the idiom portion of the present experiment, the NONLITERAL condition took somewhat longer than the PARAPHRASE or LITERAL conditions, although when an ANOVA followed by Tukey tests was performed on the idiom portion alone, these differences were not significant. It is possible that the idioms selected for this experiment included more that were unfamiliar to subjects than had been the case in Ortony et al's experiments, or alternatively, that the subject population was less knowledgeable as to their use. The overall conclusion from comparing results across experiments is that the high variability produced by using idioms and metaphorical phrases does not effectively elucidate what is occurring in nonliteral language processing. In outlining the hypotheses underlying Ortony's et al.'s (1978) experiments, predictions were made concerning the effects of contextual support on understanding time for metaphorical phrases as opposed to literal ones. An inverse relationship was proposed, with longer contexts producing shorter response times. These claims are understood to be offered ceteris paribus. However, this is not always the case, since familiarity with specific usages must also play a significant part in how readily any meaning is understood. Idioms are likely to have a consistent, easily accessible meaning, while metaphorical meanings can vary from novel to commonplace. Since it is possible that familiarity exerts a differential effect over response time at least as _p_olgertu-l-a-s---tirat af the difference between matched ---  -  --  Metaphor Comprehension - 57 metaphorical and literal phrases by Ortony et al.'s definition, it is useful to emphasize the contribution of word familiarity. Accepting their argument concerning the relationship between context and comprehension at face value, the SURPRISE condition in the present experiment should take no longer than the PARAPHRASE condition, because the context is identical in both conditions, and the word or phrase that completes the SURPRISE condition is always used literally. Ortony et al. predict that metaphors can require more time for comprehension than literal phrases when there is little contextual support and can be just as fast as literal phrases when context is adequate, but such a prediction can only be made against an assumption that the classes of literal words and metaphors are not affected by any other source of variance. Clearly this is not the case for the present experiment. Despite its having a literal word target, the SURPRISE condition takes significantly longer than all other conditions for both the idiom and metaphor portions. Such evidence strongly suggests that we must consider the effects of the target as well as the context if we are to gain a realistic insight into the processing of nonliteral language. Context has a powerful shaping effect on expectations, and these expectations are thwarted in the case of SURPRISE targets. This result is consistent with the approach of Garrod and Sanford (1988), who suggested that text interpretation occurs against a constantly changing model of meaning that includes background knowledge. In thelr model, die entities mentioned in the text  Metaphor Comprehension - 58 are represented in an "explicit focus" partition of memory that becomes mapped in an "implicit focus" partition by role pointers that encode customary expectations associated with them in "scenarios". They suggest that readers will encounter difficulty processing sentences that violate these implicit assumptions. The SURPRISE condition results support this claim. Language is not always predictable, and no matter how well developed a context is, there is always a range of possible ways a sentence or idea can turn out. A long context may produce faster response times because it constrains the possibilities more, but there is no necessary relationship between degree of constraint and metaphor processing. On the basis of response time alone, it appears that manipulating relative expectancy for literal words can produce response time differences that may be even greater than the differences that metaphors produce, and does not isolate what is unique to metaphor. Kintsch and Mross (1985) have reported that thematically appropriate words embedded in a text do not prime the identification of other words (in terms of both sense activation and sense selection) unless they are also associatively related. They conclude from these data that word identification is likely to be controlled by encapsulated input modules such as those described by Fodor (1983), that would make them impervious to top-down thematic context effects. Such a claim would mean that single word metaphors would be impossible to understand, unless the intended metaphorical meaning were somehow encoded within the  Metaphor Comprehension - 59 module prior to use. If the PARAPHRASE and SURPRISE conditions had shown no difference, it is possible that an explanation like Kintsch and Mross's might account for the data. However, their results actually showed statistically significant thematic priming in 3 of their 4 experiments. Further, they claim that such effects only occur if test words follow priming words immediately, "before the thematic context has had a chance to settle upon a particular sense of the homograph priming words". Their claim then, is not that thematic context has no effect, but only that it does not determine sense activation. The present experiments are designed to investigate the full process of understanding. Even Fodor would allow that for understanding to occur, thematic context must be involved at some stage of the understanding process. Metaphors, including those in the present study, differ widely in familiarity of use. Most dead metaphors become so conventionalized that they appear as secondary definitions in dictionaries. Other metaphors, whether they arise spontaneously in conversation or are produced as the result of conscious deliberation, are recognized as highly novel. Unless they become popularized and move into the cultural mainstream, they represent unique usages. In either case, a word is being used to convey some meaning that is not its recognized literal meaning. It is important to realize that even though dead metaphors are more conventionalized, they are not any less metaphorical in a definitional_- sense -t h a rr TroVre i me t aphors. Most research on -  -  -  -  -  —  - -  Metaphor Comprehension - 60 metaphor makes no distinction on this criterion, and metaphors used experimentally appear to be more often of the dead, or highly conventionalized type. The present experiment itself confounds this distinction. Its idioms were considered to be familiar to most subjects, but its metaphors were not evaluated with respect to familiarity, and were far more often dead than novel. A statistically more powerful examination could be performed if single word dead metaphors were studied rather than idioms, and if these results could be contrasted with those of hypothetically unlexicalized novel metaphors. This observation motivated the second experiment in this series.  Metaphor Comprehension - 61  EXPERIMENT 2 - PROCESSING DEAD METAPHORS AND NOVEL METAPHORS Dead metaphors are conventionalized metaphorical usages. According to Kittay (1987) dead metaphors are ones for which the metaphorically transferred term remains the only term in the language which expresses that concept. She explains that dead metaphors often result from catachresis, where a word is deliberately adapted to fit a previously unnamed concept and then becomes the sole referent, as in the use of the word 'current' with reference to the flow of electricity. Kittay also allows that dead metaphors may derive from unique metaphorical usages that have become widespread and common over time. She regards metaphors transformed in this way as standard uses, but stops short of equating them with literal words. This is the critical question this experiment seeks to investigate. Conventionalized metaphors have often been acknowledged to be a special problem for the study of nonliteral language processing. Dascal (1987), perhaps the strongest recent proponent for the existence of literal language, has allowed that indirect speech acts, idioms, and 'frozen' metaphors can be considered literal as much by virtue of their conventionality as their compositionality. He regards ironic utterances and novel metaphors to provide a more convincing test of whether literal meaning plays a part in nonliteral understanding. In particular, he cites  ^•:, experimen s as failing to recognize that  Metaphor Comprehension - 62 highly conventionalized usages may produce results that are indistinguishable from those of literal expressions. At the same time, however, Dascal still claims that literal meaning is always accessed regardless of context. Searle (1979), while allowing both that rejection of literal meaning takes place only if an utterance is "radically defective" and that dead metaphors eventually become literal, still contends that "Sam is a pig", when used with a human referent, would be an example of a radically defective utterance if taken literally. Accessing and rejecting the literal meaning of 'pig' would appear to be a necessary step before a metaphorical understanding is constructed in his model. Overall, there seems to be considerable confusion as to how conventionality would affect predictions concerning metaphor. Since dead metaphors represent the sole or most familiar word for a concept, but not the sole or most familiar usage of a word, they resemble both literal words and metaphors. However, in their role as metaphors, they should evoke the same cognitive processing demands as any other metaphor, unless their lexicalization is sufficient to transform them into literal words. Such a finding would certainly be problematic for threestage models like Clark and Lucy's (1975) or Searle's (1979), and would require that any research on metaphor understanding at least direct some attention toward establishing criteria for lexicalization. Experiment 2 was designed to investigate whether novel  Metaphor Comprehension - 63 metaphors would take longer to understand than dead metaphors, and, if so, whether this result could be taken as an indication that the novel metaphorical meanings had not yet become lexicalized. Three alternative results were possible. If neither dead nor novel metaphors took longer than literal words, it could be assumed that there was simply no difference between understanding literal and nonliteral words when preceded by adequate context. Finding this result would lend support to the conclusions of Ortony, et al. (1978). If both took longer, one explanation could be that all nonliteral understanding entails a distinct process of metaphorical interpretation, as the threestage models such as Clark and Lucy's (1975) seem to imply. And if dead metaphors took no longer to understand than literal words but novel metaphors did take longer, lexicalization of a specific usage as a result of familiarity, rather than any special attribute of metaphor understanding, would appear to be operative. Blank (1988) reported evidence consistent with such a hypothesis in his study. As in the previous experiment, intermediate length contexts of 1 - 2 sentences were used, and the same contextual effects as were examined in Experiment 1 were also assessed here, for the same reasons. It is essential to keep in mind that the literal targets in these experiments were words that were judged unequivocally to be conventionalized, and that what we define as novel metaphors are based on singular uses of exactly the same words. I have defined conventionalization to be a function of a consistent meaning  Metaphor Comprehension - 64 being attached to a word and not of the word itself, so that a highly conventionalized meaning and an unconventionalized meaning can both be represented by the same word, with an array of possibilities in between, that only context can clarify. The former is what has been designated a literal usage and the latter is called a novel metaphorical usage. The LITERAL and NONLITERAL conditions of this experiment represent these two possibilities, where the word is the same and the meaning is different. The PARAPHRASE and NONLITERAL conditions hold meaning constant while using different words. The use of single word metaphors in both portions makes comparing the dead and novel metaphor portions more meaningful than in Experiment 1. Reading times for the targets should be comparable, so any response time differences between the NONLITERAL conditions in the two portions of this experiment can be attributed to conventionalization differences between the two kinds of metaphors. To test whether all metaphors require essential processing that goes beyond conventionalization, the novel metaphor portion stimuli of the earlier experiment were edited to select as much as possible only those metaphors that were novel at least to the degree that they were not listed as secondary definitions of the literal word, based on Webster's New World Dictionary of American English (Neufeldt, ed., 1988). Conversely, dead metaphors were selected if they were so listed. An example of a dead metaphor item is provided in Figure 3.  Metaphor Comprehension - 65  CONTEXT  Whenever Toby gets a new mystery novel by her favourite author, she puts everything else aside because she finds it SO  ENDINGS  [NONLITERAL] absorbing. [PARAPHRASE] interesting. [SURPRISE] difficult.  LITERAL CONTEXT The major oil companies are interested in developing a material for cleaning up oil spills that is harmless, yet highly oil  ENDING  [LITERAL] absorbing.  Figure 3: Sample dead metaphor item from Experiment 2.  Metaphor Comprehension - 66 In developing items for this study, it was interesting to discover that most truly unique metaphors were not suitable because they were impossible to express in a single word for the paraphrase conditions. Metaphors that were too obscure were also excluded, but for a different reason: it was felt they could simply be impossible for subjects to understand in a reasonable length of time. As a result, Experiment 2 is a weaker test of the possible differences in processing time than might theoretically be possible if only absolutely unique metaphors had been included. In effect, this experiment only compares highly conventionalized metaphors to less conventionalized ones. Method  Subjects Thirty-two new undergraduate subjects were recruited from the University of British Columbia psychology student subject pool. Each received one class credit for an hour of participation. All claimed to be competent readers and speakers of English. Three subjects in this group were dropped and replaced, one because a low post-test score (66%) combined with very fast response times indicated that the subject was probably not reading the stimuli to the point of understanding, and two others after a preliminary analysis, because their means in over half the conditions exceeded the group means by more than two standard deviations.  Stimuli and Procedure Stimuli in the NONLITERAL condition of this experiment were  Metaphor Comprehension - 67 all chosen to be dead metaphors or novel metaphors. This essentially amounted to a partitioning of the metaphor condition of Experiment 1, with some refinement and a number of new items created. Twenty stimuli were developed for each portion, and the same four conditions as in Experiment 1 were created for each stimulus. Stimuli and targets used in this experiment and a copy of the post-test are provided in Appendix B. Dead metaphor and novel metaphor items were constructed in the same way, with similar length contexts of one or two sentences. Contexts varied in length from 15 to 28 words in the case of the dead metaphors, and 18 to 29 words for novel metaphors. Targets, with a single exception, were all single words. The design, apparatus, and presentation used in this experiment were identical to those of Experiment 1. Every subject received 160 trials of alternating blocks of 20 dead metaphor items or 20 novel metaphor items in various orders within a single session. Results and Discussion All subjects obtained scores ranging from 78-100% on the post-test, with the mean for this group falling at 93.6%. The successful retention of test items was apparently the result of subjects correctly following directions to read items to the point of understanding. Data were edited before analysis and any response time differing by more than two standard deviations from the inaiviauai subject's mean in that condition was deleted, and the  Metaphor Comprehension - 68 mean recalculated based on the remaining response times. Any response of less than 100 msec or greater than 5,000 msec was also excluded. This amounted to less than 5% of total responses. Analyses were performed both on the full range of experimental data and then just on the three base conditions, as explained in Experiment 1. Order and even/odd ANOVAs are reported for only the three base conditions. The 2 X 4 repeated measures analysis of variance showed statistically significant effects for both portion [F(1,31) = 4.22, p = .048] and condition [F(1.4,42.7) = 27.24, p < .001] and the interaction between them [F (1.7,54) = 16.24, p < .001]. These results were substantiated in the 2 X 3 analysis [portion: F(1,31)= 12.54, p = .001; condition: F(1.2,36.3)= 17.98, p < .001] and a significant portion by condition interaction [F(1.3,40.2) = 12.82, p < .001]. Results are shown in Figure 4. The dead metaphor and novel metaphor portions show similar patterns of response times except for the NONLITERAL condition, where the two metaphor conditions differed in familiarity. These results are important in that most research on metaphor seeks to identify differences between literal and nonliteral language without clearly indicating any criteria for selecting nonliteral samples, yet there is a dramatic difference between dead and novel metaphors that would necessarily influence the conclusions that can be drawn. The dead metaphors used in this experiment produced response times similar to literal words having the same ittaning (NONLITERAL [deadj,  765  msec; PARAPHRASE, 780 msec),  Metaphor Comprehension - 69  MEANS - EXPERIMENT 2  1100  1000  900 0 CO CC 800 A.  700  D NOVEL 111 DEAD N^P^L^S CONDITIONS - EXPERIMENT 2  Figure 4: Results of Experiment 2 - Dead and novel metaphor portions. (N) nonliteral [dead/novel metaphor] (P) paraphrase (L) literal (S) surprise.  Metaphor Comprehension - 70 while the novel portion shows that novel metaphors required about as much processing time as literal words whose meanings were surprising (NONLITERAL [novel], 936 msec; SURPRISE, 940 msec). The final important contrast to observe is between the NONLITERAL condition results in the two portions. Mean response time to NONLITERAL [dead] was 765 msec and to NONLITERAL [novel] was 936 msec, a difference of 171 msec. By way of contrast, the two PARAPHRASE results differ by only 25 msec and the two LITERAL results are only 7 msec apart. The statistically significant result for portion can be accounted for by the difference in familiarity between the two types of metaphor. These results disconfirm the idea that a special cognitive process is involved in metaphor processing, and point in the direction of lexicalization as the most important factor contributing to response time delays in nonliteral language understanding. When the stimuli are broken down into three groups to investigate the effects of previous exposure to context, there was no order by portion by condition interaction [F(1.8,54.9) = .77, p = .453], indicating that priming is not having a differential effect on response times in the various conditions. Stimuli divided into arbitrary groups according to context number were analyzed by comparing the results of even to odd numbered contexts. No statistically significant interaction of portion by condition by context number group was found [F(2,62) = 2.25, p = .114] in this analysis. Given these results, it appears that at least two distinct  Metaphor Comprehension - 71 factors could be affecting response times. The first factor is the degree to which the context is permitting the subject to anticipate a target meaning, as revealed by comparing the PARAPHRASE to the SURPRISE conditions. The fact that the anticipation should be the same until the point where the target is displayed indicates that it is at least partially a quality of the target that determines how quickly it is understood. This quality is not simply determined by whether a word is literal or not, since these two conditions both have literal endings. It must be the compatibility of meaning between context and target that is critical. In the present results, this advantage operates in favour of the PARAPHRASE targets. An equally important result to emerge from the present experiment is that novel metaphors take longer than their paraphrases to process. From this evidence, it is reasonable to suspect that some kind of additional processing is going on when a word with a previously lexicalized literal meaning is used metaphorically in a novel way. However, the lack of response time difference between the dead metaphors and their paraphrases, suggests that any conclusions that could be drawn about metaphor processing would be inadequate without defining exactly how conventionalized all the metaphors being investigated might be. It was felt that it would be important to do an experiment based on some objective criteria for conventionalization, both for literal and nonliteral words.  Metaphor Comprehension - 72  EXPERIMENT 3 - THE EFFECTS OF CONVENTIONALIZATION Experiment 2 provided evidence concerning the important role that conventionalization plays in the comprehension of metaphors. While it revealed that subjects respond differently to dead and novel metaphors as a class, controls for the conventionalization of any of the other targets were not imposed. As in previously published experiments investigating literal/nonliteral differences, randomness with regard to degree of conventionalization was assumed, but if there were a coincidentally higher overall level of conventionalization in the targets of one condition than in another, such influences could produce spurious effects that would seriously affect the empirical findings. A case in point might be if the PARAPHRASE words were more highly conventionalized as a class than the SURPRISE words. If so, the SURPRISE condition results could be slower for this reason alone, rather than because they disrupted the anticipation generated by the context. Since all of the previously drawn conclusions could be challenged by this possibility, Experiment 3 was designed to specifically address the question of conventionalization. If we define the mental lexicon as the words and word meanings any individual has access to as a result of his/her personal experiences with a language, it is virtually impossible to obtain information about exactly how conventionalized any  Metaphor Comprehension - 73 specific words are for any individual subject. Ideally, this information would be necessary to support the claim that metaphor processing is determined by whether the metaphorical use of a word has become lexicalized. A practical alternative is to employ a list of count values derived from word frequency analyses. Just as any word frequency count reflects the vagaries of the particular items chosen, any individual's mental lexicon reflects the vagaries of her/his own particular linguistic exposure. Yet word frequency lists are useful in a general sense because they reveal overriding frequency trends that exist in a language, particularly at the extremes, and these can be considered to parallel the linguistic exposure received by language users. For this experiment, word frequency counts were used as an objective lexicalization criterion. Experiment 3 used some of the stimuli from Experiment 2 in a similar design but with the addition of an UNFAMILIAR condition in both the dead and novel metaphor portions. Several new items had to be added in order to accommodate all of the endings. The UNFAMILIAR condition consisted of the same context that was used in the items in each of the other appropriate conditions (recall that the LITERAL condition used a different context) but ended with a literal word that was relatively unfamiliar, as determined by its frequency of use in everyday English according to Francis and Kucera's Frequency Analyses of English Usage: Lexicon and Grammar (1982). Selection of target words was aided by the Longman Synonym Dictionary (1986) and Webster's New World  Metaphor Comprehension - 74  Dictionary (Neufeldt [ed.], 1988). Care was taken to select unfamiliar words that were not so unlikely that they would not be recognized or understood by an undergraduate subject; the results could obviously have been manipulated by choosing totally obscure or archaic unfamiliar words but this was not the intent of the experiment, and such alternatives were judiciously avoided. Targets in the other conditions were refined so that they would be as equal in word frequency as possible within a stimulus item, and distinctly higher in frequency than the UNFAMILIAR targets. There is a history of empirical evidence in support of the 'word frequency effect' (e.g., Rubenstein, Garfield & Millikan, 1970). These experiments have all shown that subjects reliably respond faster to high frequency words than low frequency words. The justification for adding the UNFAMILIAR condition was to see if similar word frequency differences would emerge in this paradigm. For example, if differences in response time between the UNFAMILIAR and PARAPHRASE conditions were found to be larger than differences between the NONLITERAL and PARAPHRASE conditions, then attributing any processing time advantage to the mere fact of a word being 'literal' becomes increasingly problematic. Further, if latencies in the above comparisons are similar, then it is worth considering that similar factors might be affecting them. Such a finding would cast further doubt on the existence of a distinct metaphorical processing stage in lexical comprehension.  Metaphor Comprehension - 75 Method Subjects A new group of 40 undergraduates from the University of British Columbia psychology student subject pool served as subjects in this experiment. Each received one hour of class credit for participation. All claimed to be competent readers and speakers of English. Two of the original subjects in this group were dropped and replaced, one because of a high proportion of mean response times that exceeded the combined subject means by more than two standard deviations, and one because of a score of 50% on the post-test. Stimuli and Procedure The stimuli in this experiment were of the same type as used in the previous experiments, with the addition of the UNFAMILIAR condition. While a precise matching on word frequency for all of the other targets was obviously impossible, overall frequency counts were adjusted so that they were similar for the four replicated conditions, and much lower for the UNFAMILIAR condition within each stimulus item. Average word frequency counts in the dead metaphor portion as obtained from Francis and Kucera (1982) were: NONLITERAL, 57.75; PARAPHRASE, 54.6; LITERAL, 57.75, SURPRISE, 54.3; UNFAMILIAR, 1.35. In the metaphor portion, they were: NONLITERAL, 30.5; PARAPHRASE, 34.4; LITERAL, 30.5; SURPRISE, 33.9; UNFAMILIAR, 1.7. To recapitulate, the NONLITERAL, PARAPHRASE, SURPRISE and UNFAMILIAR conditions all n en ica con ext. The targets of the  Metaphor Comprehension - 76 NONLITERAL, PARAPHRASE and UNFAMILIAR conditions also all had meanings that were similar, but the target for the NONLITERAL condition was a metaphor (dead or novel), for the PARAPHRASE condition it was a literal word that was used with approximately equal frequency to the nonliteral word (as used in its literal sense), and for the UNFAMILIAR condition it was a literal word that was used much less frequently in normal discourse. The SURPRISE condition presented the same context, but its targets had a different meaning than the others, and one that was less likely to be anticipated, although its usage was of approximately equal frequency on the basis of the Francis and Kucera (1982) data. Very low frequency surprise words were not used, in an effort to keep the UNFAMILIAR and SURPRISE categories as distinct as possible. As before, the LITERAL condition had different contexts but the same target word as the NONLITERAL condition, but with the target word used literally. Consequently, word frequency counts in these two conditions were always identical. An example of a novel metaphor item from this experiment is shown in Figure 5. A complete list of stimuli and targets, along with a sample of the post-test, appears in Appendix C. Some of the stimulus items from the previous experiments survived intact into the present one, but many were eliminated or refined in order to bring them into line with the stricter criterion of similar word usage frequency. Stimulus sets consisted of twenty dead metaphor and twenty novel metaphor each having five variations, for a total of 100 of each,  Metaphor Comprehension - 77  CONTEXT The politician's presentation was smooth, but the facts he presented were not consistent with what Terry knew to be true. She suspected the data had been  ENDINGS [NONLITERAL] cooked. [PARAPHRASE] changed. [SURPRISE] stolen. [UNFAMILIAR] contrived.  LITERAL CON TEXT  Marilyn spent the morning in the orchard, picking apples for applesauce. She didn't stop working until evening, when the apples were finally peeled and  ENDING [LITERAL] cooked.  Figure 5: Sample novel metaphor item from Experiment 3.  Metaphor Comprehension - 78 or 200 presentations per subject. Sets were combined as before, with the additional condition making each set of 20 consist of 4 samples of each of the five conditions. Dead metaphor and novel metaphor sets were alternated, with each subject receiving a unique order of presentation of the sets. As before, a 32 item, true/false pencil-and-paper recall quiz followed the response time portion of the experiment. Presentation of stimuli, response procedure, and collection and analysis of data were all performed identically to the method described in Experiment 1. Results and Discussion Post-test scores in this condition ranged from 81-100% correct, with a mean of 93.3%. The high retention scores indicate that items were being read to the point of understanding by all subjects. Responses differing by more than 2 standard deviations from an individual subject's mean for that condition were deleted, and the mean recalculated based on remaining responses. This resulted in the loss of about 5.5% of the total data. The results of this Experiment are displayed in Figure 6. When the results of the full experiment were subjected to analysis using a 2 X 5 repeated measures ANOVA, both portion and condition were revealed to exert statistically significant effects [portion: F(1,39) = 7.87, p = .008; condition: F(1.5,57.6) = 37.16, p < .001]. The interaction of portion and condition was not significant [F(2.5,97.4) = 2.28, p = .09].  Metaphor Comprehension - 79  MEANS - EXPERIMENT 3  1400  C) 1200  2 rz 1000 ,  0 n .  cc 800  600  D NOVEL BI DEAD  N  P^L^S^U  CONDITIONS - EXPERIMENT 3  Figure 6: Results of Experiment 3 - Dead and novel metaphor portions matched for word frequency. (N) nonliteral fdead/novell (P) paraphrase (L) literal (S) surprise (U) unfamiliar.  Metaphor Comprehension - 80 Focusing on just the three base conditions, the 2 X 3 repeated measures ANOVA revealed a statistically significant difference for portion [F(1,39) = 6.91, p = .012] and for condition [F(1.4,55.4) = 11.74, p < .001], and a statistically significant portion by condition interaction [F(1.5,57.9) = 6.56, p = .006]. The interaction in these results is attributable to the slower response times for NONLITERAL stimuli in the novel portion and the faster response times for the LITERAL stimuli in the dead portion. To test for effects of priming based on presentation order, four groups based on order were formed and response times for the first vs. last presentations of a context were compared. The portion by condition by order interaction [F(1.8,68.4) = .73, p = .470] was not statistically significant. These results may be regarded as evidence against any differential priming caused by repeated exposure to the same context. Even vs. odd numbered stimuli for the NONLITERAL, PARAPHRASE, and LITERAL conditions were also subjected to analysis of variance, to provide an indication of whether these results could be generalized to other stimuli. There was no interaction of this factor with the portion by condition interaction [F(2,77.1) = 2.37, p = .101]. This experiment indicates that subjects' previously established familiarity with the target words has a dramatic effect on their response time to understand the target. It should be remembered that the UNFAMILIAR condition targets were  Metaphor Comprehension - 81 all literal words having (approximately) the same meaning and presented following the same context as both the NONLITERAL and PARAPHRASE conditions, so these three conditions are directly comparable. There was no significant difference between the response times for the dead metaphors in the NONLITERAL condition (mean = 843 msec) and their literal counterparts in the PARAPHRASE condition (mean = 848 msec). The NONLITERAL novel metaphors (mean = 922 msec) took an average of 95 msec longer to understand than the PARAPHRASE targets (mean = 827 msec) and account for the statistically significant portion by condition interaction in the 2 X 3 analysis. UNFAMILIAR targets, however, showed response time means of 1165 msec in the dead metaphor portion and 1214 msec in the novel metaphor portion. The fact that this difference overshadows the effect of the metaphorical word suggests that what we define as nonliteral is being confounded with familiarity in response time experiments. That is, if we depend on response time differences to distinguish literal from nonliteral usages without controlling for how familiar the words are, the differences we find might also be accounted for in terms of the unfamiliarity of the metaphorical usage rather than any distinct metaphorical quality. The fact that both NONLITERAL conditions are faster than their corresponding UNFAMILIAR conditions suggests that metaphor may have a facilitory function in understanding. The argument could be made that metaphors are functional in this respect, and the wi  n^nl y vo iced  observation that it is often new information that  Metaphor Comprehension - 82 is conveyed through metaphor is given additional credence by these results. The SURPRISE condition results are also important. Because this condition has the same context as the PARAPHRASE condition and is also completed by a literal word, a relatively pure measure of the effect of the compatibility of context with target word is possible when their respective response times are compared, now that word frequency has been controlled. The PARAPHRASE condition can be regarded as a baseline, since its target is a common literal word that fulfils the conventional expectations generated by the context. It should give the fastest response times of any of the conditions sharing the same context. According to the data, the effect of thwarting contextual expectations in the SURPRISE condition slows responding twice as much on average as having a novel metaphorical target does: the difference between the SURPRISE and PARAPHRASE response times was 192 msec, while the difference between NONLITERAL and PARAPHRASE response times was 95 msec. The difference between the SURPRISE and PARAPHRASE conditions shows that response times can be different despite having identical contexts and literal word endings of equal frequency. This difference may be attributed to context-target compatibility. A word that is highly predictable within the context is easily understood, while a word that is less predictable, though no less grammatical, tends to disrupt understanding to a marxed aegree. Since literal words have shown  Metaphor Comprehension - 83 response time differences that are both shorter (for paraphrases) and longer (for unfamiliar words) than those for nonliteral words, it is not tenable to hypothesize an additional processing stage to account for the understanding of metaphors, at least until we can isolate the effects of familiarity. On the other hand, the entire responsibility for determining rapid understanding cannot simply be attributed to a compatible context. The PARAPHRASE and UNFAMILIAR conditions had the same context and the same meaning, and yet produced different response times. If context were the dominant source of meaning, neither metaphors nor unusual or unfamiliar words should show any response time differences. The slower times for the UNFAMILIAR when compared to the PARAPHRASE conditions make it plain that understanding is at least partly determined by the familiarity of the target word itself. Nor is it correct to say that a metaphorical target categorically takes longer to understand than a literal one, since these results show the opposite when the literal target is surprising or unfamiliar. Even stronger evidence for the contribution of familiarity is provided by contrasting the results of the SURPRISE and UNFAMILIAR conditions. Both have identical contexts and literal word endings. The difference between them is that the context supports the meaning of the UNFAMILIAR targets but not the SURPRISE targets, and this lack of contextual support has been shown in the two previous experiments to cause response slowing. By any 'compatible context plus literal word' hypothesis that  Metaphor Comprehension - 84 ignores familiarity, the UNFAMILIAR targets should be faster than the SURPRISE targets, and yet they are not. The fact that the SURPRISE endings have higher word frequency counts than the UNFAMILIAR endings accounts for the difference, regardless of the UNFAMILIAR condition's contextual compatibility advantage.  Metaphor Comprehension - 85  EXPERIMENT 4 - VOLITIONAL METAPHOR GENERATION If dead metaphors can be understood as quickly as literal words, can they also be produced just as readily? Is there a difference in the capacity to generate novel as opposed to dead metaphors? This experiment looks at metaphor from the standpoint of volitional generation in response to a context. The evidence of the three previous experiments suggests that novel metaphors are unlexicalized uses of lexicalized words. It seems unlikely that we would have independent mental lexicons for comprehension and for generation, but if we generate speech just from the words whose meanings we have lexicalized, where do metaphors (or more precisely, novel metaphors) come from? The implication from the previous experiments in this thesis is that novel metaphor generation should come only from a linguistic need that is not represented by a previously lexicalized word. Since the word chosen has to have enough common aspects that the new meaning can be made comprehensible through its use, it would seem that such a selection would need to be carefully made, and should take longer than generating highly lexicalized words. Schank's (1982), Glucksberg's (1991), and Chandler's (1991) models all explain metaphor comprehension by an expansion of the meaning of the vehicle to incorporate the topic. Cutting across theories somewhat, we could say that in generating a novel -^• •^-^• - n  1  y a wor•as a thematic relationship that  Metaphor Comprehension - 86 could be applied to another concept, and label the theme with the concept prototype from which it arose. This would mean that the metaphorical meaning did not exist unless it was needed. Attention in this fourth experiment is directed toward illuminating the question of the lexicalization of metaphors. When subjects are given the opportunity to freely generate any kind of sentence endings, do they respond by accessing lexicalized words first? If so, would dead metaphors be among those so accessed? Do the endings they can generate parallel word frequency information - i.e., are the first endings subjects generate high frequency words, with lower frequency words and novel metaphors coming afterwards? Is it even possible to generate novel metaphors on demand? In this experiment, subjects were asked to generate their own single word endings for some of the same sentence contexts as were used in the previous experiments to see what kinds of endings would be generated, how consistent they would be between subjects, whether the order in which words were generated would correspond to printed word frequency data (as an indication of lexicalization), and whether novel metaphors would occur at all amongst the generated data. Method Subjects  Sixteen undergraduate psychology students from the University of British Columbia participated in the experiment for one hour of class credit. These students had not participated in  Metaphor Comprehension - 87 any previous experiment in this series. Stimuli and Procedure Subjects were given an instruction sheet to read and then were handed a booklet consisting of 20 brief (1-2 sentence) contexts, in which the last word of the last sentence was missing. Three contexts were presented per page, and subjects were told to work on one until a buzzer sounded as a signal to move on to the next. Two minutes were allowed for each context item. A countdown was displayed on a video screen beside the subject, to help them keep track of what context number they were on. The contexts were a selected group of stimulus items from Experiment 3, presented verbatim without targets. Each context was followed by a total of 12 blank lines. The first 10 lines were intended for subjects to write as many suitable endings to the context as occurred to them, and the last two were for endings that might suggest themselves while working on another question, after the time interval was over. Subjects were cautioned not to fill up spaces with synonyms of a single idea, and flexibility of approach, including the possibility of metaphors, was specifically encouraged. Responses were screened for appropriateness, grammaticality and metaphoricity by the author and by an independent coder, who then collaborated to resolve discrepancies. All contexts and tabulated responses appear in Appendix D.  Metaphor Comprehension - 88 Results and Discussion Subjects were able to come up with a surprising number of plausible possibilities in response to the stimuli, averaging more than 6 per context per subject. Further, there was little intersubject overlap, resulting in a large total number of different responses. Overall, 79% of raw responses were unique to a single subject. Even after editing to remove ungrammatical responses, the minimum number of responses generated by the 16 subjects in any context was 57 and the maximum was 162. The nature of the contexts was such that most responses were adjectives describing some aspect of a person or situation mentioned in the contexts. The overwhelming majority of responses (92%) were words used literally. The definition of literal word usage was based on the intuitive judgement of the two coders, who proved to have been highly consistent. Virtually all of the remaining responses that could be classified as metaphors (8%) were the most conventionalized kind of dead metaphor (e.g., brilliant, cold, distant). Figure 7 illustrates a typical context taken from the experiment and the number and variety of words produced in response to it. Within each context, words that occurred a single subject had a significantly lower printed word frequency (Kucera and Francis, 1968) than did words that occurred to more than one subject [F(1,19) = 6.73, p = .018]. The data from this analysis are illustrated in Figure 8. The first word that occurred to a subject in a specific  Metaphor Comprehension - 89  Figure 7: Sample item from Experiment 4, with responses.  STIMULUS CONTEXT AND EDITED RESPONSES Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were response [#of Ss]  ^  ** = metaphor  28. implemented [1] 1. acceptable [1] 29. important [1] 2. adequate [2] 30. indifferent [1] 3. appropriate [1] 31. inevitable [1] 4. approved [1] 32. inflexible [1]** 5. bad[1] 33. intelligent [1] 6. [the] best [1] 34. irreversible [1] 7. changing [1] 8. continuing [1] 35. justifiable [1] 36. justified [1] 9. correct [4] 10. critical [1] 37. law [1] 38. long-range [1] 11. decided [2] 12. definite [1] 39. made [2] 40. minimal [1] 13. demanded [1] 14. different [1] 41. necessary [3] 42. normal [1] 15. established [1] 43. official [1] 16. excellent [1] 17. fair [1] 44. overdue [1] 18. final [2] 45. permanent [1] 19. finalized [1] 46. positive [1] 20. finished [1] 47. practical [1] 21. firm [2]** 48. proceeding [1] 22. fixed [1] 49. progressing [1] 23. foolproof [1] 50. proven [1] 24. good [2] 51. reasonable [3] 25. great [1] 52. responsible [1] 26. ignored [1] ^53. responsive [1] 27. immutable [1] 54. right [3]  55. righteous [1] 56. rigid [1]** 57. settled [1]** 58. silly [1] 59. substantial [1] 60. superior [1] 61. supportive [1] 62. transferred [1] 63. unchangeable [2] 64. unstoppable [1] 65. untransferable [1] 66. useful [1] 67. useless [1] 68. wanted [1] 69. well-considered [1] 70. wise [1] 71. workable [1] 72. working [1]  Metaphor Comprehension - 90 context was also more likely to occur to other subjects than the last word was [F(1,19) = 28.65, p = .001]. These data are shown in Figure 9. On the other hand, there was no indication that these first words had a higher printed frequency than the last words [F(1,19) = .63, p = .438]. These results support the belief that novel metaphors represent unlexicalized uses of previously lexicalized words. The absence of any but the most conventionalized metaphors suggests that producing novel metaphors is very likely not achieved through direct lexical access. It appears that a task of this type encourages subjects to search their personal lexicon for suitable word endings, and novel metaphors are not present in that lexicon. Individually, the first completion that occurs to most subjects is often the same. Later, the possible ways of construing the context tend to diverge, but the frequency ratings of the words selected does not decline. This suggests that the lexicon is not organized as a single dictionary, but rather that it may be parsed into discrete units based on semantic similarity (Jastrzembski, 1981). Subjects appear to be moving from unit to unit rather than scanning an entire list, whether ordered alphabetically or by overall frequency, although frequency information may play a part within a unit. These results are also revealing in terms of the contribution of context to understanding. None of the contexts was tightly constraining to subjects, despite the fact that only the last  Metaphor Comprehension - 91  100 90 >0  z  0  80  LL  a 70 cc 0 60 2 50  40  ONE^ONE SUBJECTS MAKING SAME RESPONSE  Figure 8: Results of Experiment 4 - Comparison of the printed word frequency of endings generated by only one subject in a context with the number generated by more than one subject.  Metaphor Comprehension - 92  6 Lu  z  0 a. 5 cc  2 4 U)  0  z  3  2 z  40  1  FIRST^LAST ORDER OF OCCURENCE OF RESPONSE  Figure 9: Results of Experiment 4 - Comparison of mean number of subjects making the same response for first generated and last generated words.  Metaphor Comprehension -  93  word of the last sentence was missing. While the finding that first words were more likely to be consistent between subjects than last words were suggests that there is likely some culturally controlled tendency to construe contexts similarly, the imbalance of different word endings over repeated ones testifies to an amazing flexibility and openness in language that further argues against understanding dominated by contextual constraint.  Metaphor Comprehension - 94  EXPERIMENT 5 - SEMANTIC DECISION Response latency to single word endings proved useful as a measure of comprehension for the first three experiments, precisely because it produced the confounding of familiarity with metaphoricity that led to attempts to isolate these factors in the later experiments. At the same time, however, it presented a serious problem. Since the presentation subjects were given consisted of the context followed by a single word target and required only a keypress when the target was understood, subjects could hypothetically keypress their way through the entire experiment although their understanding of the targets might be variable, vague, or nonexistent. The obtained results could be a consequence of such a tactic, since longer hesitations would be more likely in the case of the more unusual targets in the NONLITERAL, SURPRISE, and UNFAMILIAR conditions regardless of whether they were ultimately understood. Even the post-test might not ensure that the results were valid, as many questions could be answered by recalling the meaning of any of the various words that completed the context. Because of these possible sources of error, and despite the fact there has been a wellestablished precedent for using simple response time in experiments of this type (Hoffman and Kemper, 1987), it was felt that the conclusions that could be drawn from these results would be more valid if they could be replicated using a different  Metaphor Comprehension - 95 paradigm. Experiment 5 was intended to replicate Experiment 3, using choice response time to a semantic decision task. Hoffman and Kemper (1987) have pointed out that time to understand metaphors may not be a reliable indication of what is actually going on when experimental subjects are presented with artificially created metaphors. Nonetheless, in any well designed experiment, consistent differences must offer evidence of some psychological reality, and the refinements produced by employing choice response time potentially provide more information than simple response time does. A further advantage of the choice task is that it offers some insight into where and how the delays that influence response time arise. Specifically, the data from this experiment were analyzed by counting the number of errors in each condition, and separating these from the responses dropped because of response time violations (i.e., longer than two standard deviations beyond the individual subject's mean in each condition or less than 100 msec or more than 10,000 msec). Doing this provides an additional perspective on why some conditions take longer than others. For example, UNFAMILIAR words might be expected to produce a bipolarity of response depending on whether a target is recognized at all. This would indicate that if a word is known, it presents no problem, but if it is not known, subjects are simply unable to make use of it, and must resort to what they can extract from context, resulting in a cluster of longer times on some of the items. An error  Metaphor Comprehension - 96 analysis should also prove enlightening about what happens when metaphors are encountered. In Experiment 3, the SURPRISE and UNFAMILIAR words took longer in both portions than NONLITERAL words. Does this mean that the metaphors are more efficiently conveying their meaning, or is it that they are perceived as so incongruous that the subject considers them nonsense, and simply responds randomly? Discovering whether subjects make more errors in one condition than another would help answer this question. Method Subjects Twenty new subjects completed this experiment. Most were undergraduate students at the University of British Columbia who participated for class credit in psychology, while the remainder were paid volunteers. Six of the original subjects were dropped and replaced due to combined error and response delay deletions exceeding 25/200, or 12.5% of total responses. In addition, a pre-experiment was run to validate the choice of target and non-target items. Five subjects completed this portion, but two clearly misunderstood a large proportion of the items as well as the task. These respondents gave evidence that English was not their first language, and their results were not taken into consideration as validation data. Stimuli and Procedure In this experiment, a semantic decision task was employed where subjects had to make a choice between an appropriate and an inappropriate target with respect to meaning. Subjects were  Metaphor Comprehension - 97 presented with a 1 - 2 sentence context, displayed all at once on a video screen. Once they felt they had read and understood the context, they were to press the bottom buttons on the response box. This removed the context and presented an appropriate and an inappropriate target word, side by side, on the screen. They were to choose the word that made the most sense in terms of the context they had just read, and to indicate their choice by pressing a button on the right or left, corresponding to the position on the screen of the correct target word. Stimulus contexts were taken from Experiment 3. Each target was paired with a nontarget word of identical length and closely matched printed word frequency count, as determined from Kucera and Francis (1968). Nontargets were also selected for grammatical appropriateness to the sentence context. All stimuli and targets used in this experiment are included in Appendix E, along with a copy of the post-test. A pre-experiment was undertaken in which three subjects were run through the entire experimental procedure, but rather than responding as quickly as possible to their choice of a target word, they were asked to rate each word, target and nontarget, on a seven point scale for comprehensibility and appropriateness; that is, whether each word made sense, given the context. A rating of 1 indicated 'highly appropriate' and a rating of 7 indicated 'highly inappropriate'. Any items for which a target was given a rating of 4 or more, or a nontarget was given a rating of 4 or less, or the two words were given ratings less  Metaphor Comprehension - 98 than 4 scale values apart, were identified and replaced in order to establish a clear distinction between targets before the main part of the experiment was run. For the actual experiment, the presentation of sentence targets proceeded as in Experiment 3, using identical equipment, except that two words appeared on the screen side by side following the removal of the context, rather than just one. Subjects used a four button response box with the thumbs of both hands pressing the two bottom buttons simultaneously to present the target word choices after the context had been read and understood, and the index fingers of each hand pressing the top button on the right or left corresponding to the choice of the correct target. This posture was designed to standardize response time, eliminating the variability introduced by subjects moving their hands onto the button before pressing it. The twothumb technique was also intended to prevent a hand bias from influencing the target response. Choice of side for occurrence of each correct target was randomly selected for each subject within the presentation program. Choosing a target word automatically resulted in the presentation of the next stimulus context in the set. Stimuli were presented in sets of 20, with the order of presentation of sets randomized between subjects, except that dead metaphor and novel metaphor sets always alternated, just as in Experiment 3. Following the stimulus presentation portion of the experiment, subjects were given a 32-item true/false pencil-  Metaphor Comprehension - 99 and-paper quiz, as a means of ensuring that they had actually read and understood the stimulus items. Results and Discussion Scores on the post-test ranged from 81-100% correct, with a mean of 93.5%. These results strongly indicated that subjects experienced no difficulty in correctly reading or understanding the stimuli. Data were screened to eliminate incorrect responses and any times under 100 msec or over 10,000 msec. Based on the remaining data, all response times greater than 2 standard deviations from an individual subject's mean for a given condition were also dropped and the mean recalculated. Any subject for whom this screening procedure resulted in the elimination of more than 25 responses (12.5% of their data) was dropped from the analysis and replaced by a new subject. For subjects remaining in the analysis, data dropped by all three screening processes represented 381 responses in total, or 9.5% of all data - 3.8% because of decision errors and 5.8% because of time delays. A portion by condition (2 X 5) analysis of variance of response times for the whole experiment revealed main effects for both portion [F(1,19) = 10.9, p = .004] and condition [F(2.1,40.7) = 49.37, p < .001], as well as a statistically significant interaction between the two factors [F(2.4,45.8) = 11.86, p < .001]. When only the NONLITERAL, PARAPHRASE, and LITERAL conditions were analyzed, the results were consistent with the five condition analysis (portion: [F(1,19) = 13.42, p =  Metaphor Comprehension - 100 .002]; condition: [F(1.3,24.1) = 26.1, p < .001]; interaction: [F(1.2,22.8) = 21.72, p < .001]). Results are shown in Figure 10. Clearly, the portion by condition interaction arose from the difference in response times between the two portions in the NONLITERAL condition. The order in which a context was viewed (first vs. last exposure) did not interact significantly with the portion by condition interaction [F(1.3,25.2) = 1.8, p = .192]. When stimuli were divided into two arbitrary even/odd numbered halves and compared as a measure of generalizability, the half by portion by condition interaction was not statistically significant [F(2,38) = 2.14, p = .132]. These analyses may be interpreted to indicate that priming effects are not responsible for the portion by condition interaction, and that the characteristics of the stimuli are not idiosyncratic. In reaction time experiments, Posner (1978) has distinguished the faster times recorded in simple as opposed to choice reaction time as the difference between automatic activation and conscious attention. This was not intended to be the case in these studies, as any understanding task necessarily requires conscious attention, and reading the preceding context decisively separated Experiment 3 from studies where responding was based simply on detection. However, the necessity of reading two words on the screen as opposed to one, and of actually making a decision as to which was correct, was expected to lengthen response times in this experiment in contrast to those of  Metaphor Comprehension - 101  MEANS - EXPERIMENT 5  2000  1000  N^P^L^S^U  D NOVEL ■ DEAD  CONDITIONS - EXPERIMENT 5  Figure 10: Results of Experiment 5 - Dead and novel portions in a semantic •^I I  I I I  (S)surprise (U)unfamiliar.  ! ^! ^! ^'  'I •  Metaphor Comprehension - 102 Experiment 3. This effect occurred as expected, but the trends of the two experiments, with one notable exception, were much the same. This congruency lends support to the validity of conclusions drawn from the earlier experiments. The most notable change in the pattern of results of the present experiment from that of Experiment 3 occurred in the SURPRISE condition, where response times were relatively longer than in the previous experiments. This condition also produced substantially more errors than any of the other conditions. Counts of the total number of errors occurring in each condition of Experiments 5 and 6 are provided in Table 2. Since the error rates are assumed to represent cases where subjects have insufficient information to make a correct decision and are forced to guess, it is probably safe to assume that the number of times this happens is roughly twice the obtained error rate that is, an equal number of times when the subject makes a guess, that guess is fortuitously correct. In a count of error by condition, the SURPRISE condition showed 30 errors in the dead portion and 57 in the novel. In both portions this was more than twice as many errors as occurred in all other conditions combined. The 2 X 5 analysis of variance for errors showed a significant effect for portion [F(1,19) = 10.72, p = .004] and condition [F(2.3,44.6) = 23.84, p < .001] and a significant interaction of portion and condition [F(2.7,52.2) = 5.35, p = .004], attributable largely to the SURPRISE condition results. At the same time, the number of time  Metaphor Comprehension - 103 violations stayed fairly consistent over conditions. Given the above results, it seems that when contextual expectations are not supported, subjects become confused. With the single word targets of Experiment 3, they can simply do a mental 'double take' if necessary to make sense of what appears on the screen, and force a new meaning out of the context they had just read. This process takes time, and is probably sometimes incomplete or unsuccessful, but the single button response paradigm is insensitive to these failings, as subjects will press the button after attempting to derive meaning from the presentation regardless of how incomplete their understanding is. In the semantic decision design of this experiment, subjects encountering an ending that is different than they anticipated may attempt to test both targets in succession both before and after they revise their interpretation of the context. The high error rate may reflect the tacit time pressures, where subjects confused by having their expectations thwarted may conclude that neither target makes sense and simply respond with a random guess. In contrast, there were far fewer errors in the UNFAMILIAR conditions, about the same number of time violations, and mean response times were faster by about 80 cosec. The advantage for subjects in this condition could be that the meaning of the context was 'dependable' so the task might be simply to use whatever lexicalized information they had available about either the target or the nontarget to guide them to a correct choice more often. No 'double-take' would be  Metaphor Comprehension -  104  TABLE 2. ERROR AND TIME VIOLATIONS: EXPERIMENTS ^5 and 6 EXP 5^EXP 6  EXP 5  EXP 6  PORTION  CONDITION  DEAD  NONLITERAL  8  12  24  25  DEAD  PARAPHRASE  6  7  25  26  DEAD  LITERAL  1  2  22  25  DEAD  SURPRISE  30  38  24  24  DEAD  UNFAMILIAR  14  11  24  26  NOVEL  NONLITERAL  12  51  25  21  NOVEL  PARAPHRASE  4  4  21  22  NOVEL  LITERAL  6  20  17  21  NOVEL  SURPRISE  57  33  27  23  NOVEL  UNFAMILIAR  13  24  21  22  ERRORS  TIME  Metaphor Comprehension - 105 advantageous, and the longer response times could simply be reflecting time to summon more incidental clues, such as morphemic information. The combination of longer response times and higher error rates in the SURPRISE conditions draws attention to the critical role played by context in developing consistent expectations. The suggestion that the expectations of context exert strong constraints on understanding might be challenged by citing the results of Experiment 4, where it was shown that contexts were interpreted in any number of different ways. However, it is important to remember that in Experiment 4 subjects were explicitly instructed to use the contexts flexibly, and not to paraphrase the same idea in different words. This might have required them to duplicate the 'double-take' reading of the context each time. In any event, the results of this experiment can not be taken as weakening the conclusions to be drawn from that one, or vice versa. This result provides new information regarding the importance of context in comprehension which the previous experiments may have disguised. Its full impact, in light of the results of Experiments 3 and 4, will be discussed later. Briefly, it can simply be said that any less anticipated ending, whether NONLITERAL, SURPRISE, or UNFAMILIAR, will produce some delay, but the fact that the meaning expectations of the NONLITERAL and UNFAMILIAR conditions remain consistent helps these words to be understood more effectively than the SURPRISE  Metaphor Comprehension - 106 condition, which requires a revision of how the context is interpreted to be understood. One of the most interesting findings of this experiment is that understanding due to the presence of a metaphorical target is clearly not as much of a handicap to achieving understanding as either having an unanticipated or an unfamiliar target. In both portions, the NONLITERAL conditions produced fewer errors than the UNFAMILIAR and SURPRISE conditions, and no more time violations. Further, 14 of 20 subjects made no errors at all on the novel metaphors and 13 of 20 made no errors on the dead metaphors. From this vantage point, it begins to look as if metaphorical usages of conventionalized words might actually provide a cognitive advantage over either the SURPRISE or UNFAMILIAR conditions, by allowing the use of the more familiar meaning of the metaphorical word to assist the hearer, particularly if the information to be conveyed is new (surprising) or incomprehensible (unfamiliar). To further investigate whether metaphor actually contributes to understanding, and to gain further insight into why the SURPRISE condition of Experiment 5 produced such a high rate of errors relative to all of the other conditions, it seemed useful to see what might happen if context were made a less useful component of understanding. If contextual information were reduced to a bare minimum, would the SURPRISE target simply become impossible to interpret. Might the lack of well developed contextual  Metaphor Comprehension - 107 expectations hinder the comprehension of novel metaphors? Ortony et al.'s (1978) work with long and short contexts appeared to point in this direction. It was in pursuit of answers to these questions that Experiment 6 was developed.  Metaphor Comprehension - 108  EXPERIMENT 6 - REDUCED CONTEXT Experiment 6 focuses attention on the role of context in understanding. In Experiment 5, the delay in understanding time caused by a SURPRISE target exhibited itself in both longer response times and higher rates of errors than in any of the other conditions. The SURPRISE condition responses took even longer than the UNFAMILIAR condition responses, which is a reversal of direction from the Experiment 3 results. Given this turn, it now seemed appropriate to delve more deeply into the component of contextual preparedness, both to verify the intuitions raised by the results of the previous experiment and to further explore their implications. The presentation for this experiment was identical to Experiment 5, but contexts were rewritten to greatly decrease their length and informational content. Subjects in the two experiments were matched by presentation order. It was expected that reduced contexts would still provide sufficient information for rapid processing of PARAPHRASE and LITERAL conditions, but that the UNFAMILIAR and SURPRISE conditions would suffer - the former because the context was less able to create a full meaning that would allow a less familiar word to be recognized as informative, and the latter because the lack of contextual support combined with decreased contextual information would make the choice between targets particularly difficult to resolve.  Metaphor Comprehension - 109 Since the SURPRISE condition would be more confusing, errors were expected to be high; the UNFAMILIAR condition could just show longer response times. Results in the NONLITERAL condition were less predictable from prior results. If the novel metaphors maintained their advantage over the SURPRISE and UNFAMILIAR conditions when context was reduced, this finding would support the hypothesis regarding the functional advantage of novel metaphors put forth in Experiment 5. The NONLITERAL condition in the dead metaphor portion was again predicted to produce results similar to its PARAPHRASE counterpart. Decreasing contextual information would be expected to affect both the PARAPHRASE and LITERAL conditions similarly, by lengthening response time and increasing errors, but to a lesser degree than in the other conditions. The attempt to reduce the length of the stimulus contexts had parallels with Ortony et al.'s (1978) experiment where metaphor understanding was measured in long and short context conditions. In their experiment, the short contexts consisted of just the first sentence or phrase of their 5 - 6 sentence long contexts, followed by the target. The example offered in their paper, with short contexts underlined here, is as follows (p. 467) : Literal inducing context: Approaching the enemy infantry, the men were worried about touching off landmines. They were very anxious that their presence would be detected prematurely. These fears were compounded by the knowledge that they might be isolated from their reinforcements. The outlook was grim.  Metaphor Comprehension - 110 Metaphorical inducing context: The children continued to annoy their babysitter. She told the little boys she would not tolerate any more bad behavior. Climbing all over the furniture was not allowed. She threatened to spank them if they continued to stomp, run, and scream around the room. The children knew that her spankings hurt. Target: Regardless of the danger, the troops,(sic) marched on. Given the example provided, one can see that the strategy of limiting a presentation to just the short context plus target does more than just reduce information. The reader could be left with virtual nonsense. In the literal example, the semantic conjunction of the words 'enemy infantry' in the context with 'danger', and 'troops' in the target is likely enough to trigger a response indicating that the target was understood, although even prolonged contemplation of the presentation would yield little real understanding. For the metaphorical presentation, no such conjunction occurs and the lack of coherence is manifest. It is hardly surprising that understanding time was slowed when contexts were shorter. The intention underlying the present experiment was to reduce context information, but to maintain enough relevant cues so that targets could be clearly differentiated. Rather than simply processing a context until enough information is gleaned so that a metaphor makes sense, it seems more reasonable to claim that the reader is continually involved -rpre a ion process, invo ving e ements of  Metaphor Comprehension - 111 both context and target. Finding that metaphors are not slowed when context is shortened and the quality degraded would cast doubt on Ortony et al.'s findings with regard to the relationship between context length and metaphor, suggesting instead that Ortony's et al.'s shortened contexts slowed metaphor understanding because they lacked sufficient comprehensibility, rather than that were simply shorter in length. The length of contexts in both experiments was roughly equal, averaging 6 words in Ortony et al's study and 6.8 words in the present one. The alternative possibility is that when contextual support is reduced, the NONLITERAL novel metaphors would produce the same kind of confusion as the SURPRISE condition showed in Experiment 5, resulting in the same consequences of longer response times and higher error rates. Method Subjects A new group of twenty subjects from the University of British Columbia subject pool participated in this experiment for class credit. All indicated a fluency in English. Two subjects whose error rates alone exceeded 15% and whose post-test results were 75% or less were excluded from analysis. A third subject was dropped because of response latencies that exceeded the means of all other subjects by two standard deviations in more than half of the conditions, and another was eliminated when she revealed that she had participated in one of the earlier experiments in this series. Four replacement subjects provided  Metaphor Comprehension - 112 satisfactory data. Stimuli and Procedure The stimuli used in this experiment were all derived directly from the stimuli of Experiment 5, and are presented in Appendix F, along with a copy of the post-test. They were written with the intention of reducing both the length and information content of the contexts as much as possible, while preserving enough of the gist to make the targets comprehensible. Most proper nouns were changed to pronouns, unless multiple pronouns made reference impossible to determine. Dead metaphor contexts were reduced from a mean length of 23 words as they appeared in Experiment 5 to between 3 and 12 words (mean = 6.58). Novel metaphor contexts were reduced from a mean length of 25 words to between 3 and 10 words (mean = 7.08). With a few slight exceptions to preserve readability/grammaticality, all edited contexts contained exactly the same words in the same order as the originals. An example of an item from Experiment 6 is shown in Figure 11, along with its source item from Experiment 5 for comparison. Targets to items in Experiment 6 were identical to those of Experiment 5, as were nontargets, with a few exceptions where the reduced context made the nontarget newly plausible, and an alternative nontarget with a similar word frequency count was substituted. Apparatus and procedure were identical to those described for Experiment 5. The correspondence between the stimuli of Experiment 5 and those of Experiment 6 provides an opportunity to compare these  Metaphor Comprehension - 113  EXPERIMENT 5 CONTEXT Edmund had written two novels before he turned 30. Later he wrote three more, but his editors rejected them all. They complained that his work had become  EXPERIMENT 6 CONTEXT They complained his work had become  ENDINGS  [NONLITERAL] stagnant tapestry. [PARAPHRASE] dull sand. [SURPRISE] childish tropical. [UNFAMILIAR] hackneyed laminated.  EXPERIMENT 5 LITERAL CONTEXT It was shocking for Arnie to revisit his old home town after 20 years away. Ugly townhouses replaced open fields, and the pond where he often swam was now  EXPERIMENT 6 LITERAL CONTEXT The pond where he swam was now  ENDING [LITERAL] stagnant perverse. Figure 11. Sample novel metaphor item from Experiment 5 and its counterpart from Experiment 6.  Metaphor Comprehension - 114 experiments to see exactly how much difference the context manipulation made. Although the two experiments were run on different groups of subjects, the order of presentation in both was matched between groups, and a number of between-groups analyses were conducted. The results and discussion in this section is limited to Experiment 6. An evaluation of the implications of the two experiments taken together will be found in the following section. The Experiment 5 post-test was edited and items truncated to eliminate information no longer contained in the stimuli. This made them somewhat vague, and it was anticipated that scoring would demand more leniency than in previous experiments. Results and Discussion Scores on the post-test ranged from 63-94% correct, with a mean of 81.05%. Due to the vagueness of some of the questions, subjects scoring at the low end of the scale were retained if their data were otherwise acceptable. The overall analysis of variance for five conditions showed statistically significant effects for portion [F(1,19) = 63.66, p < .001] and condition [F(3.3,62.2) = 44.81, p < .001], and a significant portion by condition interaction [F(3.8,72.6) = 11.51, p < .001]. The three condition ANOVA produced identical levels of significance to those for the five conditions [portion: F(1,19) = 83.75, p < .001; condition: F(2,38) = 20.04, p < .001; interaction: F(1.5,28.3) = 22.66, p < .001]. The results of Experiment 6 are shown in rigure 13. Again, the NONLITERAL metaphors in the novel  Metaphor Comprehension - 115  MEANS - EXPERIMENT 6  2000  1500  O NOVEL 1000  111 DEAD  N^P^L^S^U CONDITIONS - EXPERIMENT 6  Figure 12: Results of Experiment 6 - Dead and novel portions in a semantic decision task with reduced context. (N)nonliteral [dead/novel metaphor] (P)paraphrase (L)literal (S)surprise (U)unfamiliar.  Metaphor Comprehension - 116 portion were responded to more slowly than their PARAPHRASE counterparts, while the dead metaphor portion produced about the same response latencies in both the NONLITERAL and PARAPHRASE conditions. Looking at the effects of first vs. last (fourth) presentation of the same stimulus, it was found that there was no order by portion by condition interaction [F(1.6,31) = 1.29, p = .286], indicating that repeated presentations of the same context were not priming one condition more than another. The effect of dividing the stimuli into two groups based on whether arbitrarily assigned item numbers (but having the same assignment as the items in Experiment 5) were odd or even produced a statistically significant interaction of arbitrary half with portion and condition [F(1.6,30.8) = 8.87, p = .002]. Such results, along with those of the order analysis, indirectly argue in favor of the repeated measures design of the present studies, since priming effects were found to be negligible, while item differences and intersubject differences can be substantial. An error count was made to determine if reducing context had affected one condition more than another. Each subject received 20 presentations in each condition and 20 subjects were used, resulting in 400 presentations per condition across subjects, and 4,000 total presentations for the 10 conditions. ^Data discarded as the result of errors accounted for 5.0% and time violations accounted tor  b.8,  for a total deletion of approximately 10.9%  Metaphor Comprehension - 117 of all responses. The error and response time violation data are provided in Table 2. Error results were examined in a 2 X 5 portion by condition ANOVA, which revealed a statistically significant effect for portion [F(1,19) = 42.56, p < .001] and condition [F(3.4,64.9) = 18.33, p < .001] as well as a significant interaction of the two [F(3.4,65.2) = 11.86, p < .001]. The number of errors in the novel portion was almost twice the number in the dead portion. Post-hoc Tukey tests, with familywise error set at a = .05, showed that the NONLITERAL condition of the novel metaphor portion had significantly more errors than any other condition. Errors in this condition alone represented 25% of all errors for both portions combined. The SURPRISE condition for both portions also produced more errors and differed significantly from all but the novel portion NONLITERAL and UNFAMILIAR conditions. One completely unexpected result that emerged from this experiment was that, in the novel portion, response times for the LITERAL condition were now significantly slower than for the PARAPHRASE condition and error rates were correspondingly higher. The reverse had been the case in every previous experiment. Even that finding was somewhat curious, since both the LITERAL and PARAPHRASE conditions could be described similarly as presenting a literal word target following a compatible context. It would seem most likely that response times should be about equal for the two conditions, but if not, the expectation might be in the other direction, since the PARAPHRASE contexts had appeared  Metaphor Comprehension - 118 previously with the same meaning in both the NONLITERAL and UNFAMILIAR conditions, while the LITERAL condition context had never been seen before. Any priming effects as a result of previous exposure should favor the PARAPHRASE responses. Two possible explanations for the findings of the earlier experiments were either that it was the previous exposure to the target, rather than the context, that was responsible for priming, or that repeated exposure to the same context actually had an inhibiting effect, even when understanding was straightforward, since the context was not always 'reliable' because of the SURPRISE condition. The former explanation seems dubious, given that the target word was required to have two different interpretations in the two exposures. The latter is more plausible, although order analyses consistently demonstrated that there was no statistically significant difference in response time between first and last presentations of contexts. It is only possible to speculate, but one explanation for the earlier results is that, if metaphor processing requires the extension of the meaning of the metaphor to include the context as Glucksberg and Keysar (1990) suggest, the process would have already taken place for approximately half the number of times the target appeared in a LITERAL context (i.e., those times it appeared as a metaphor first). The literal uses would then be somewhat easier to process. The contexts required no expansion themselves, since they provided an abundance of relevant information. The difference in the present experiment is that  Metaphor Comprehension - 119 the contexts did not provide enough information to constrain the target selection, and so rather than constraining interpretation of the targets, what was required instead was an elaboration of the contexts. Such elaboration could only occur  post  hoc, that  is, after exposure to the target choices. Now the advantage is reversed, since the likelihood of the contexts already having been encountered with a target of the same meaning was 3:1 (the SURPRISE condition, while sharing the same context, would not contribute to the effect). Such an advantage would likely have been distributed over the three conditions (NONLITERAL,  PARAPHRASE, UNFAMILIAR) to the point that it was undetectable individually, but appeared as a relative handicap for LITERAL responses. Put another way, each time a LITERAL or NONLITERAL target occurred, a different context would have to be elaborated, since the meaning of the target in that context was different, while the process would only need to occur once for the 3 repeated conditions, since the meaning of the target in that context was the same. Additional support for this possibility is provided by the fact that the dead portion LITERAL results remained consistent with those of the earlier experiments, and almost identical to results in the NONLITERAL condition (1138 and 1144 msec, respectively). This result is exactly what would be predicted by the argument that dead metaphors represent a case in which a class-inclusion had been established previously, through repeated exposure so the same broadened definition is available to both usages. The high rate of errors in the SURPRISE  Metaphor Comprehension - 120 condition is also consistent with this explanation, since it also represents a case where a unique meaning must be understood. The hypothesis that novel metaphors actually promote understanding when context is reduced is not supported by these results of this experiment. The dead metaphors in the NONLITERAL condition were not notably different from the PARAPHRASE results in terms of response times (1144 msec and 1178 msec, respectively) although errors were somewhat higher. In contrast, the response times to novel metaphors in the NONLITERAL condition were now even slightly longer than SURPRISE condition responses (1588 vs. 1549 msec) and over 400 msec longer than the comparable PARAPHRASE condition (1182 msec). Further the error rate for novel NONLITERAL responses was dramatically higher when contrasted to the PARAPHRASE condition (51 vs. 4). In fact, 12.75% of all responses to novel metaphors were wrong choices. Suggesting that metaphor provided additional information from which to construct a complete meaning seems plausible until the error results of Experiment 6 are considered. What could account for such a result? The first, and most obvious option is simply that Ortony et al. (1978) were right - that subjects need a long, well developed context to understand metaphors. In this case, it is worthwhile to point to the NONLITERAL results in the dead metaphor portion and emphasize that even if Ortony et al. are correct, such a finding apparently only applies to novel metaphors, since the two portions behave in quite different ways. A  second option is to consider these results as somehow  Metaphor Comprehension - 121 idiosyncratic to a few subjects. However, in Experiment 6 every subject made at least one error in the NONLITERAL condition, and 6 of 20 made 4 or more, while in Experiment 5, 14 of 20 subjects made no errors in this condition, and none made 4 or more. This appears to indicate a consistent effect across subjects, but such results would gain reliability if they were found to replicate in a new study, preferably with an entirely different stimulus set. It is probably true that at least some of the effects are attributable to the specific stimuli used. With these considerations in mind, a third possibility may be raised. According to Glucksberg and Keysar (1990), metaphor arises when a topic is included in a class to which it had not previously belonged, because features of a hierarchically superior category get extended from the basic level term. In the example "My job is a jail", the characteristics of a jail are extended to include my job, which in 'literal' usage would not qualify for inclusion in such a category. When presented with a very underdeveloped context, it is likely that subjects are unable to make use of the relevant metaphorical class inclusion structure, particularly when they are unsure of which of the two targets require expansion in the first place. Consider as an example one PARAPHRASE item from the experiment as shown in Figure 12. When we have read a context like "They complained his work had become" and the target choices are "sand" and "dull", we can make a choice even though context information is limited. The only content word in the context is "work" and we are far better able  Metaphor Comprehension - 122 to develop a class inclusion for "work" with "dull" than with "sand". The choice in the NONLITERAL condition for the same context is between "tapestry" and "stagnant". The correct choice is now much less apparent. Work may have some elements in common with stagnation, but it might also be possible to find ways that work resembles tapestry. It has been said that we can find some commonality between any two things, if we try hard enough. This may be what is happening here. The struggle and ultimate guesswork involved in resolving this choice is reflected in longer response times and more errors. Now, compare the same stimulus context as it appears in Experiment 5. It reads: "Edmund had written two novels before he turned 30. Later he wrote three more, but his editors rejected them all. They complained that his work had become". Given the same two choices, the preference for "stagnant" over "tapestry" is unequivocal. We realize that the work referred to is writing novels, that rejection of later work is involved, and we can see the consistency between this contextual implication and the features of the word "stagnant" that refer to a once attractive entity that has lost its vitality, to a dullness and lack of progressive movement, to a lifelessness and lack of appeal. Instead of forcing a connection from word to context, the choice is between a word that appears patently nonsensical and one that now seems not only eminently suitable, but rich with appropriate resonances. It also seems pertinent to the discussion to think about the  Metaphor Comprehension - 123 target choices from the point of view of language use generally. The recent attention to the study of pragmatics has shown how much the derivation of meaning from language consists of contextual extrapolation. Developmental studies of language learning (e.g., Newport, Gleitman, and Gleitman, 1977) have revealed that children must constantly constrain meaning in almost miraculous ways in order to connect what is meant to what is said. In such circumstances, it is essential to be open to meaning, rather than searching for closure. At its extreme, we equate 'closed-mindedness' with 'literal' thinking and an unintelligent denseness. It is the impression we get when someone doesn't pick up on our irony, or the embarrassment we feel when we realize we have been unable to detect someone else's sarcasm toward us. The situation created when context is severely limited means that there is no certainty of closure. The results of this experiment support Ortony et al.'s (1978) conclusions in one way, but not in another. Short contexts produce longer response times, and affect the understanding of novel (but not dead) metaphors more than literal words. Understanding appears to require abstracting information from the metaphor to fit the context as much as it requires abstracting information from the context to fit the metaphor. Perhaps the most surprising finding from this experiment is that response times in the majority of conditions were no longer than those of Experiment 5. This result is elaborated on in the next section, but it is interesting to discover that even a very  Metaphor Comprehension - 124 limited context can provide enough information to permit understanding when class inclusion is unproblematic.  Metaphor Comprehension - 125  BETWEEN GROUPS ANALYSIS OF EXPERIMENTS 5 and 6 In Experiment 6, information from the Experiment 5 contexts was reduced by changing nouns to pronouns and eliminating extraneous, but schemata enhancing, details. The same orders of presentation were used in both experiments. This design resulted in the presentation of a long and short version of every stimulus that permitted a between-groups analysis with experiment (i.e., context length) as an additional factor. The comparison resembled Ortony et al.'s (1978) metaphor experiment in which different subjects received either long or short contexts for an entire experiment. Further, by combining the two experiments (Experiment 5 and Experiment 6), the total number of subjects was the same as Ortony et al.'s (40) although many more stimulus presentations per subject were made (200 as opposed to 16). It was expected that overall response times, time response errors, and target choice errors would be greater in Experiment 6 than in Experiment 5, given that contexts were reduced to approximately one-quarter their former length, and that context has consistently been recognized as an important component in linguistic understanding. However, when the response time results of Experiment 5 and Experiment 6 were compared in a 2 X 2 X 5 - experiment (5 vs. 6) by portion by condition between/within ANOVA, there was no statistically significant difference between the overall mean response times of the two  Metaphor Comprehension - 126 experiments [F(1,38) = .02, p = .878]. Although there were significant reversals in some conditions (to be discussed), the overall means for the two experiments differed by only 13 msec (1391 msec and 1404 msec, respectively). The number of deletions because of responses of either less than 100 msec, greater than 10,000 msec, or two standard deviations beyond the subject's mean for that condition were almost identical (230 in Experiment 5 and 235 in Experiment 6), but the number of response errors was higher in Experiment 6 (437 as opposed to 381). Reducing contextual information may have specific effects on understanding, but it does not appear to have had any overall effect on how long it took to choose a target and respond to it in the semantic choice task used here. Table 2 provides a comparison of response time violations and error rates for all of the conditions in the two experiments. The same ANOVA also showed that the experiment (5 vs. 6) by portion by condition interaction was only marginally statistically significant [F(3.1,117.3) = 2.38, p = .070]. Figure 13 makes this effect apparent. Of particular interest is the LITERAL condition, which was much slower when context was minimal, and the SURPRISE condition, which unlike the disadvantage it suffered in Experiment 5, was faster than the UNFAMILIAR condition in Experiment 6, and even slightly faster than NONLITERAL novel metaphors. Reducing contextual information appears to have had little effect on the processing of UNFAMILIAR  words, and to have actually sfleedkresponding imrthe SURPRISE -  Metaphor Comprehension - 127 condition for both dead and novel portions. The same pattern of results as the five condition analysis was found when just the NONLITERAL, PARAPHRASE, and LITERAL conditions were assessed. No statistically significant difference was found between the two experiments [F(1,38) = 2.09, p = .156], and the experiment by portion by condition interaction was not significant [F(1.3,50.8) = 1.09, p = .320]. Results of Experiments 5 and 6 may be compared in Figure 13. Patterns of errors between the two experiments were also examined. These are shown in Figure 14. The ANOVA for overall mean errors between Experiments 5 and 6 (long vs. short context) showed no significant difference [F(1,38) = 2.67, p = .11], but the experiment by portion by condition interaction was significant [F(3.1,119.5) = 11.54, p < .001]. The error pattern that occurred in the novel portion of Experiment 5 may account for much of this difference. It is also interesting to contrast patterns of results for mean response times with patterns for mean errors. Clearly, the information they provide is quite different. In particular, the error rates for novel metaphors in the reduced context experiment show a dramatic rise over all of the other NONLITERAL conditions, and the SURPRISE conditions for Experiment 6 shows more errors than the UNFAMILIAR conditions, even though the SURPRISE condition response times are faster. One expectation of shortening the contexts is that subjects might make greater use of extra-contextual information, including  Metaphor Comprehension - 128  MEANS - EXPERIMENT 5 AND 6  2000  1500  1000  N  P^L^S  ^  ■ NOVEL6 DEAD6 O NOVELS • DEADS  U  CONDITIONS  Figure 13: Results of Experiments 5 and 6 together - Dead and novel portions in a semantic decision tasli. (N)nonliteral (P)paraphrase (141iterat (S)surprise (U)unfamiliar.  Metaphor Comprehension - 129  ERROR MEANS - EXPERIMENT 5 & 6  3  2  0  N  P^L^S  ^  ■ NOVELS D DEAD6 El NOVEL5 ■ DEAD5  U  CONDITIONS  Figure 14: Error means for Experiments 5 and 6 together - Mean number of errors per subject for dead and novel portions in a semantic decision task. (N)nonliteral (P)paraphrase (L)literal (S)surprise (U)unfamiliar.  Metaphor Comprehension - 130 what could be gleaned and retained from previous presentations of the same stimulus. This would be most clearly demonstrated in a difference between first and last presentations in repeated contexts having a consistent meaning. However, when the means for first vs. fourth presentation for the various conditions were examined, only the UNFAMILIAR condition showed a clear trend, and this was no more pronounced in Experiment 6 than Experiment 5. Nevertheless, this effect was very large. First presentations of UNFAMILIAR targets averaged nearly 300 cosec slower than fourth presentations. The order ANOVA between Experiments 5 and 6 together for the three base conditions (with the SURPRISE and UNFAMILIAR excluded) shows no significant interaction of order with the experiment by portion by condition results [F(1.6,59.4) = .07, p = .888]. First appearances of a condition were not responded to significantly slower than fourth appearances overall. However, the even/odd analysis showed that arbitrary half did interact with the experiment by portion by condition results in the base conditions to a statistically significant extent [F(1.6,59.8) = 5.58, p = .01]. UNFAMILIAR conditions may be particularly disadvantaged since it is likely that neither target word can provide much help in understanding the meaning of the context (recall that both target and nontarget were always matched for word frequency). However, it might also be predicted that the developed context of Experiment 5 would consequently be more useful in understanding  the UNFAMILIAR targets- The _results de-not support this -  -  Metaphor Comprehension - 131 expectation. Experiments 5 and 6 do not differ appreciably in overall response time for their respective UNFAMILIAR conditions. It appears that if a word is unfamiliar, the information provided by the immediate context is not a great help in understanding it. The speculation that the LITERAL condition takes longer when context is short because contexts must be expanded post hoc to fit the targets was advanced in the discussion of the Experiment 6 results. If we accept this proposal, a neat explanation for the SURPRISE results emerges. The SURPRISE condition is the only condition similar to the LITERAL in the sense that the meaning developed by the repeated context, and consistent over three other conditions, has become inappropriate. The conditions are also similar in that they consist of a context followed by a literal word ending. The results section of Experiment 5 discusses why the SURPRISE targets would require a post hoc revision of the contexts. Since the other conditions have long contexts and do not require this procedure, the SURPRISE condition is substantially handicapped. In Experiment 6, the SURPRISE condition still takes longer than the base conditions, and accounts for a substantial portion of errors. However, relative response times are faster in Experiment 6. Possibly, the 'double take' revision of contextual meaning that took place in Experiment 5 is of little use when contextual information is already diminished. The closure that accounted for the surprise effect does not occur, and so revision of contextual expectations is less importa___14nw-il—een-te)cttrar b 5c p etions are weaker, the -  - --  -  Metaphor Comprehension - 132 subject is more open to alternative possibilities. Another way to say this is that the SURPRISE targets were simply less surprising, because shortening the contexts made subjects less rigid in their expectations. The question of why subjects could respond to stimuli just as quickly in Experiment 6 as Experiment 5 is truly puzzling, and merits further investigation. The only, highly tentative, explanation that seems plausible is that the cognitive approach called for by the lack of closure when context is reduced gives a response time advantage, independent of any advantage conveyed by solid information. This hypothesis will be further developed in the General Discussion to follow.  Metaphor Comprehension - 133  GENERAL DISCUSSION Three notable findings relating to metaphor understanding emerge from the present series of experiments. The first is that metaphor comprehension appears to be so dependent upon lexicalization - not of a word per se, but of the specific way the word is used - that it is impossible to draw conclusions about the distinction between literal and metaphorical from response time studies without simultaneously considering its effects. Variations in familiarity can reasonably account for all the response time differences found in these within-subjects metaphor experiments and can explain why, in various empirical studies, different forms of nonliteral language (idioms, metaphors) have shown differing relative response rates when compared to literal words. This is not to say that metaphorical language is in no way different from literal language, but only that it is probably not different in terms of demanding a discrete processing stage where inappropriate literal meanings are tested against context and subsequently rejected and revised. While these experiments refute the implication of stage models that every metaphor should show a longer response time, they have demonstrated that a temporally expensive process of matching meaning to context is taking place for unfamiliar words, or familiar words used in a surprising or unfamiliar way, with novel metaphors_being---a—eliaracter i g ti a example of the latter. - -  -  - -  —  Metaphor Comprehension - 134 On the other end of the familiarity continuum, dead metaphors can be as familiar as so-called literal words, and can typically be accessed just as quickly. The lexicalized metaphor is still a metaphor and as such represents a prototypical exemplar of a class. The difference is that the development and expansion of the ad hoc class it represents is not required since dead metaphors, like literal words, have categorical definitions that are already lexicalized. In terms of language, this means that multiple uses are known and available to the user. Both may be considered exemplars of different aspects of the same meaning, and they would likely be represented in a dictionary as primary and secondary meanings of a single word. The second result of note involves the contribution of contextual preparedness to comprehension. Context is a very significant factor in shaping expectations for understanding what is to follow. In the present study, targets with meanings supported by context (in the PARAPHRASE condition) were consistently processed more quickly than targets with meanings unsupported by context (in the SURPRISE condition). Since targets in both these conditions were literal words matched for familiarity in an identical context, the differences in processing time must have to do with the fact that the meaning of the SURPRISE targets is less likely to be anticipated. The ability of context to constrain possible alternative meanings is both a strength and a weakness. As a strategy it is parsimonious tot_e_abla----te—anticipate toih at as to come. Such an —  Metaphor Comprehension - 135 approach enhances both speed and depth of understanding. At the same time, we could virtually never absorb anything new if we were always able to anticipate what we are about to hear. Context would be detrimental if it constrained language to such a degree. Experiment 4 shows the error of expecting that meaning can be fully constrained by context, even when that context provides a fair amount of information. The final two experiments reveal that overall comprehension speed is not affected by a less developed context, but that certainty of understanding may be weaker. Perhaps metaphor intervenes in the comprehension process by forcing us to remain flexible. We are particularly inclined to reserve judgment when we encounter a novel metaphor in a situation where context has not provided sufficient constraint since we might be wrong - but we know we might be wrong, and this strategy allows us to make a graceful recovery as context becomes elaborated. To illustrate, consider the metaphor "Juliet is the sun" first cited by Searle (1979) and also employed by Chandler (1991). This phrase is impossible to understand in isolation and without reflection. If we have minimal context (say, what we can remember about Shakespeare's Romeo and Juliet - and assuming that we have even had such exposure) we might be able to make some sense of it if we can recognize it was spoken by Romeo, who was infatuated with Juliet. Reading the play and encountering the line as Romeo alludes to the sun at daybreak, when the moon becomes pale by__colaparlson r--we can laelin to see ways that the sun -  -  -  Metaphor Comprehension - 136 has features that Romeo could apply to Juliet. When we fully examine the play, we find that it is permeated with light and cosmology imagery, providing a considerable body of context. However, it is arguably really only Romeo (or Shakespeare) who can truly understand what he means when he utters the metaphor because it is only he who can know what he has encoded in the class of sensations he associates with the sun. If we are to make progress toward ecological validity in metaphor research, we must recognize that for a subject in a psychology experiment to paraphrase what a metaphor means, and for a psychologist to evaluate such a description, moves well away from directly accessing the cognitive processes of language understanding. The third area into which the preceding experiments have been able to provide insight concerns the question of how the target words worked together with context to create meaning. Experiments 1, 2, 3, 5, and 6 each had stimuli in which conditions had identical contexts but different endings. Among these, novel metaphors took less time to understand than surprising or unfamiliar words, and dead metaphors were virtually no different than literal words once word frequency was controlled. A particular strength of the class-inclusion model is that it provides a reason why metaphor produces this advantage. When we are processing new information, metaphor permits that information to be expressed in a way that we can access through what we already know. Through the mutually compatible processes of -earrstrairit from context and enhancement  Metaphor Comprehension - 137 from metaphor, the likelihood of successful communication is greatly increased. The particular suitability of using metaphor for providing new information raises an important point about feature-matching or 'topic and vehicle' theories in general. As Glucksberg and Keysar (1990) point out, theories that depend upon matching features of the topic with those of the vehicle, whether by interaction (Black, 1979), salience imbalance (Ortony, 1979), similarity (Tversky, 1977), common ground (Verbrugge and McCarrell, 1977) or interactive competition (Chandler, 1991) assume that the features of the topic are as well known as those of the vehicle. If this were true, metaphor would only be redundant, not informative. Saying "Mary is a kitten" if we know that Mary is charming and playful only confirms what is already known. Saying "Mary is a kitten" if we don't know anything at all could very well mean that we are discussing a household pet. It is only if there is an intent to inform - if we know Mary is a human and we are discussing her appeal - that the metaphor becomes useful. This imbalance between metaphor as the source of information and topic as the recipient of it is important, and has not been formally developed. Metaphors provide or confirm features about the topic, overwriting default values for underdeveloped schemas with specific information. For this reason, all successful metaphors implicitly depend upon some sort of pre-established contextual framework. Returning_totne_T.Juliet-is ^e sunit example, feature -  -  Metaphor Comprehension - 138 matching theories would find difficulty with this metaphor, since there is very little interaction on a feature-by-feature analysis that could yield a foundation on which to provide information. When Glucksberg and Keysar (1990) claimed that metaphorical comparisons are semantically asymmetrical and fundamentally irreversible, the explanation would seem to be because the features of the metaphor relevant to the information being conveyed are always better known than those of the topic. This is why we must know a literal meaning of a word in order for us to understand the metaphor derived from it. Such a conclusion supports the contention that literal meaning is fundamentally related to familiarity with a specific usage and represents what I believe literal meaning to be. Glucksberg and Keysar's (1990) attempt to draw a distinction between what is literal and what is metaphorical runs into a problem which I believe the class-inclusion model can successfully explain. They claim that one clear difference between the two is the availability of the class-inclusion construction for metaphoric, but not literal, comparisons. Metaphoric statements, which compare two unlike things, can be paraphrased to look like class-inclusion statements (My job is like a jail/ My job is a jail), while literal statements, which compare two like things, can't be paraphrased in this way (Bees are like hornets/ *Bees are hornets). The problem with this contention is that there would appear to be just as many literal statements that_am_r—lass—ineittsloTi ttder, but they can't be --  Metaphor Comprehension - 139 paraphrased as similes (Bees are insects/ *Bees are like insects). The explanation for this effect does not lie in how alike the things are, but rather in the fact that all classinclusion statements involve categories at different levels, while all similes involve categories at the same level. Because the class-inclusion model uses the same metaphorical word as both type and token, the metaphor can be referred to on either level. That is, if we say "My job is like a jail", we are using "jail" as a category at the same level as "my job". If we say "My job is a jail" we are using "jail" as a prototypical exemplar (and de facto name) of a superordinate category. It might still be reasonable to state that only metaphor has this linguistic flexibility and that in contrast, literal statements may only be parallel or hierarchical, but never both. It is easiest to think of a metaphor as providing new information, but we can easily think of examples where this is not the case. If we both know Mary equally well, and we both know that she is charming and playful, then it could be argued that saying "Mary is a kitten" is not providing new information. However, if such a statement were uttered, its intent would likely be to elicit confirmation of the speaker's observation with a second opinion. This is new information of a somewhat different kind, but it also illustrates a useful linguistic function fulfilling a common human need. Research on idiom and proverb understanding (Hoffman and _ Kemper, 19871 has_shown-that priming with one proverb helps in  Metaphor Comprehension - 140 the recall of another with the same meaning, and that an initial focus on the image of the proverb can detract from comprehension of the implicit figurative meaning, while an initial focus on the meaning makes the images subsequently more meaningful. This abstraction of meaning is consistent with Bartlett's (1932) findings about subjects retaining the gist of stories, while adapting the details to previously formed expectations. This seems to be essentially what is happening on a different scale in the case of semantic priming, when more abstract categorical similarities are recognized. If the same process happens with a word when it is used metaphorically, the gist of the word in its literal sense provides a bootstrap in the form of a categorical inclusion into an abstract, thematic class which can make comprehension more complete. No feature matching is necessary. Only the 'literal' meaning of the metaphor needs to be jointly understood. The similarity between Glucksberg's (1991) hierarchically superior metaphorical class, Shank's (1982) abstract, thematically structured TOPs, and Rosch's (1973; 1978) superordinate level categories is readily apparent. If a TOPlike structure can explain semantic linkages between stories and experiences, it can also explain why concepts like 'doctor' and 'hospital' can be semantically linked, and why we can employ a similar process in understanding what is meant by "My job is a jail". Some recent challenges to Glucksberg and Keysar's (1990) class-inclusion model lave---come froffi Gibbs (1992). Gibbs's -  Metaphor Comprehension - 141 criticisms focus on the belief that metaphor is a fundamental characteristic of how people categorize and make sense of their experience. Drawing on Lakoff's (1987) ideas concerning image schemas, Gibbs asserts that the process of understanding metaphor using ad hoc superordinate categories is rare. More usually, he contends, metaphors derive from concepts already existing in long-term memory. He cites "Love is a journey" as an example of a fundamental structural metaphor. While Gibbs's assertion that metaphor usage reflects a fundamental process related to categorization and meaning may be correct, Glucksberg, Keysar and McGlone (1992) have produced empirical evidence that subjects interpreting metaphors are not constrained by basic image schemas. It seems unlikely that something like the "love is a journey" metaphor is inherently fundamental, for many of the same reasons why any proposal of innate concepts seems problematic. We can certainly understand the meaning of a metaphor couched in an infinite number of completely independent conceptual frameworks; i.e., "Love is a rose" or "Love is a warm puppy". Virtually all investigations of metaphor in some way allude to its very ability to restructure familiar things in creative ways - essentially a definition of the process of ad hoc categorization. Systematicity in metaphor, in my view, arises more from the conventionalization process (a particularly apt metaphor becomes culturally widespread and then generates variants on a consistent theme) than from innate or even cultural constraints- Thc-falyt that th-éCiass-inclusion -  Metaphor Comprehension - 142 model specifies that the conceptual class framework accessed by a metaphor is previously unnamed provides a model for how the interaction between thought and language is achieved. Lakoff's model would appear to depend upon foundational metaphors being rigidly lexicalized, an idea that the preceding experiments appear to refute. Although the model presented by Glucksberg and Keysar (1990, Glucksberg, 1991) provides what seems to me the most viable explanation of metaphor understanding to date, and one that is supported by the data of the experiments presented here, there are some points of their theory that in my opinion could be strengthened by elaborations suggested by the present studies. Glucksberg and Keysar (1990, Glucksberg, 1991) decisively reject a three stage model like Clark and Lucy's (1975) (i.e., the Standard Pragmatic Model), citing the evidence of Ortony et al. (1978), Gibbs (1984), and others who have shown that metaphors in context need take no longer to understand than comparable literal sentences. The reader is left with the implication that a classinclusion model would predict that metaphors and their literal counterparts are processed equally quickly. However, in Ortony et al.'s experiments, the finding of 'no difference' occurred for idioms in one experiment, and metaphorical phrases in elaborated contexts in another. In most of Gibbs's experiments, idioms or indirect requests were studied and then generalized to explain all nonliteral language. Both bodies of work focus on uses of nonliteral language-that-----from the point of view of the present  Metaphor Comprehension - 143 studies appear to be special cases. While the present results support the class-inclusion model as an explanation of the process involved in understanding metaphor, it seems consistent to suggest that when the process of expanding a superordinate class in order to encompass a new member is required, such a process takes more time than when it is unnecessary - that is, when the class into which the inclusion takes place is already named and lexicalized. In advancing this explanation I am not suggesting that a different sort of three-stage model is called for. We do not always need to test and expand a hierarchical class when any metaphor is encountered, as the dead metaphor results consistently show. What I am saying is that there is a continuum of latency depending on whether or not we need to 'enlarge our understanding' as it were, and the enlarging process takes time. The time required appears to be related to how familiar we are with the meaning of the word used metaphorically. The second qualification to the work of Glucksberg and Keysar (1990) comes from the importance of context to understanding, and to understanding metaphors in particular. Context is not simply a matter of a single topic word for a metaphorical vehicle. The hypothetical "a is b" metaphor virtually never exists in isolation. It is important to recognize that context is always working to constrain possibilities, so much so that a very well developed context will likely result in_am interpretatiorviar metaphor by default, as  Metaphor Comprehension - 144 Ortony et al. (1978) suggest. A tentative model for a theory of metaphor understanding based on a combination of context and word or phrase target can be described by a lock and key analogy, with context represented by the lock and the various target possibilities by the keys. To illustrate how this model would work, consider the sample stimulus item from the novel metaphor portion of Experiment 5 presented in Figure 12. In the context of a promising young writer whose works are now being rejected, the PARAPHRASE target "dull" is well supported by the context, although any of a number of other choices with different meanings could fit as well, such as repetitive, boring, or dated. In the analogy, the context has provided a clearly defined but undifferentiated keyhole that is so generic that even if the intended key were missing, we could find another one that would fit adequately. However, if the context were to become very highly developed, the keyhole would become reciprocally more constrained, finally reaching the point where only a single key would be likely to fit. When we have a context that shapes the target to this extent, we can recognize the appropriate key even in its absence, simply by the shape of the keyhole. This anticipatory advantage makes the search process highly efficient. In the case of the UNFAMILIAR target, "hackneyed" is a word with a similar meaning that is usually understood but less often used. It fits the context, but takes longer to process because it is less familiar— We-knew-whwt itloOks like, but we have to -  Metaphor Comprehension - 145 search the key ring before we can locate it, and some of us might search for quite a while but be unable to find it. The SURPRISE target "childish" also takes longer than the PARAPHRASE, but for a different reason. It has a well lexicalized meaning, but that meaning does not immediately appear to fit the context. The SURPRISE meaning is a strong key that reshapes the context keyhole to make itself fit. The same schemata are present, but different parts are involved. We focus less on the fact that the author was successful in the past, and more on his success being tied to his youth. Reshaping the keyhole also takes time. The lexical meaning of the NONLITERAL target "stagnant" seems not to fit the keyhole at all, but one aspect of it fits the same portion of the context as does "dull". Abstractly, both "dull" and "stagnant" can be members of a class that might be characterized as static or uninteresting. In addition, however, a rich web of other associations in the class typified by stagnant (but not by dull) like stale, unmoving, and overripe, along with nonverbal sensory ones, enrich our understanding of how the editors regarded the author's work. The metaphorical key not only fills the keyhole completely, it overfills it, expanding the context to take it all in. Some of the questions raised by the Gibbs (1984, 1989) Dascal (1987 - 1989) debate over whether literal meanings exist at all are clarified by the results obtained in these studies. Briefly, Gibbs claimed that literal meanings do not have psychological reality -s4nee-meaning is not simply a function of  Metaphor Comprehension - 146 compositional analysis and since his results have shown that a literal analysis stage is not obligatory in the determination of meaning. Dascal argued that literal meaning can be equivalent to conventionally established meaning, and in this sense, literal meaning both does exist and does participate in the determination of speaker meaning. By equating what is ordinarily defined as literal meaning with conventionality, both author's points are accommodated. Within the framework of literal and metaphorical that he employs, Gibbs is correct in saying that literal language is not always accessed before a nonliteral interpretation is made, as his own results with idioms show. On the other hand, consideration of novel metaphors confirms Dascal's point regarding the need to determine meaning when a usage is not conventionalized. Then the literal meaning of the word must play a role. The longer response times to novel metaphors in the present experiments demonstrate the effects of this process. Perhaps the simplest way to sum up the findings of the studies developed for this dissertation is to say that they present metaphor as much less mysterious than many conceptions convey. If we see metaphors as words or phrases that mean something completely different than we want to say, they seem a bizarre, irrational aberration of language. If instead we see metaphors as words or phrases that convey new information in a way that makes it more easily comprehensible, they seem a flexible, efficient tool of language. This approach to metaphor is remarkably consisterit-with-the AzMy we understand other ---  -  Metaphor Comprehension - 147 cognitive processes, like learning, memory, and concept formation to operate. While the actual processes involved in categorization and generalization that permit this to happen remain speculative, the results of these studies unequivocally point toward the functional view.  Metaphor Comprehension - 148 CONCLUSION The present experiments have shown that three factors: familiarity of usage, contextual preparedness, and conceptual class, represented by the literal meaning of words used metaphorically, all contribute to understanding any discourse. Familiarity effects can account for the processing delays associated with nonliteral language. Since identifying nonliteralness with unfamiliarity of usage can explain metaphorical effects, the question of what possible function metaphor might have then occurs. In reply, I have proposed that when context is adequate but the concept is new, complex, unique, or superlative, using metaphor provides a way of approaching the idea through class-inclusion. By employing what is known about a familiar class to provide informational depth to a new one, metaphor is given a functional explanation. When context is inadequate, metaphors are much less useful but possibly serve to alert us to seek extracontextual cues. The time cost of understanding metaphors is compensated for by the quality of information they can convey. If metaphors provide us access to information that would otherwise be unknown and inexpressible, the additional time it takes to understand them seems a small price to pay.  Metaphor Comprehension - 149 REFERENCES ALLMAN, W.F. (1989). Apprentices of Wonder: Inside the Neural  Network Revolution. New York: Bantam Books. ALLEN, P.A., MCNEAL, M, & KVAK, D. (1992). Perhaps the lexicon is coded as a function of word frequency. Journal of  Memory and Language, 31, 826-844. BARTLETT, F.C. (1932). Remembering: A Study in Social  Psychology. London: Cambridge University Press. BLACK, M. (1979). More about metaphor. In A. Ortony (Ed.)  Metaphor and Thought. New York: Cambridge University Press. BLANK, G.D. (1988). Metaphors in the lexicon. Metaphor and  Symbolic Activity, 3(1), 21-36. BOBROW, S. & BELL, S. (1973). On catching on to idiomatic expressions. Memory and Cognition, 1, 343-346. CHANDLER, S.R. (1991). Metaphor comprehension: A connectionist approach to implications for the mental lexicon. Metaphor  and Symbolic Activity, 6(4), 227-258. CHOMSKY, N. (1965). Aspects of the Theory of Syntax. Cambridge: M.I.T. Press. CLARK, H.H. & LUCY, P. (1975). Understanding what is meant from what is said: A study in conversationally conveyed requests.  Journal of Verbal Learning and Verbal Behaviour, 14, 56-72. DAMASIO, A.R. & DAMASIO, H. (1992) Brain and language. Scientific  American, 267(3), 88-95. DASCAL M. (1987). Defending Literal Meaning. Cognitive  Science,_11, 259-241.  Metaphor Comprehension - 150 DASCAL, M. (1989). On the Roles of Context and Literal Meaning in Understanding. Cognitive Science, 13, 253-257. FODOR, J.A. (1975). The Language of Thought. Cambridge, MA: Harvard University Press. FODOR, J.A. (1983). The Modularity of Mind. Cambridge, MA: MIT Press. FORSTER, K.I. (1990). Lexical processing. In D.N. Osherson & H. Lasnik (Eds.) Language: An invitation to cognitive science, Vol. 1. Cambridge, MA: The MIT Press. FRANCIS, W.N. & KUCERA, H. (1982). Frequency Analyses of  English Usage: Lexicon and Grammar. Boston: Houghton Mifflin. GARROD, S. & SANFORD, T. (1988). Thematic subjecthood and cognitive constraints on discourse structure. Journal of  Pragmatics, 12, 519-534. GIBBS, R.W. (1979). Contextual effects in understanding  indirect requests. Discourse Processes, 2, 1-10. GIBBS, R.W. (1980). Spilling the beans on understanding and  memory for idioms in conversation. Memory and Cognition, 8, 449-456. GIBBS, R.W. (1984). Literal meaning and psychological theory.  Cognitive Science. 8, 275-304. GIBBS, R.W. (1989). Understanding and literal meaning.  Cognitive Science. 13, 243-252. GIBBS, R.W. (1992). Categorization and metaphor understanding. Psychological Pev ew- ^X52=-511.  Metaphor Comprehension - 151 GLUCKSBERG, S. (1991). Beyond literal meanings: The psychology of allusion. Psychological Science, 2(3), 146-157. GLUCKSBERG, S. & KEYSAR, B. (1990). Understanding metaphorical comparisons: Beyond similarity. Psychological Review, 97(1), 3-18. GLUCKSBERG, S., KEYSAR, B., & McGLONE, M.S. (1992). Metaphor understanding and accessing conceptual schema: Reply to Gibbs (1992). Psychological Review, 99(3), 578-581. GRICE, H.P. (1975). Logic and conversation. In P. Cole & J.L. Morgan (Eds.) Syntax and Semantics: Speech acts. Vol 3. New York: Academic Press. HEIDER [ROSCH], E. (1972). Universals in color naming and memory. Journal of Experimental Psychology, 93, 10-20. HOFFMAN, R.R. & KEMPER, S. (1987). What could reaction time studies be telling us about metaphor comprehension? Metaphor and Symbolic Activity, 2(3), 149-186. HUBBELL, J.A. (1988). The effects of metaphorical and literal processing on subsequent lexical latencies of sentence components. Ames, IA: Unpublished doctoral dissertation. JASTRZEMBSKI, J.E. (1981). Multiple meanings, number of related meanings, frequency of occurrence, and the lexicon. Cognitive Psychology, 13, 278-305. KINTSCH, W. & MROSS, E.F. (1985). Context effects in word interpretation. Journal of Memory and Language, 24, 336-349.  KIRK, R.E. (1982). Experimental Design: Procedures for the Behavioral Sciences. Monterey, CA: Brooks/Cole.  Metaphor Comprehension - 152 KITTAY, E.F. (1987). Metaphor: Its Cognitive Force and Linguistic  Structure. Oxford: Clarendon Press. KUCERA, H. & FRANCIS, W.N. (1967). Compositional Analysis of  Present Day American English. Providence, RI: Brown University Press. LAKOFF, G. (1987). Women, Fire, and Dangerous Things: What  Categories Reveal About the Mind. Chicago: University of Chicago Press. LAKOFF, G. & JOHNSON, M. (1980). Metaphors we Live By. Chicago: The University of Chicago Press.  Longman Synonym Dictionary. (1986). Essex: Longman Group, U.K. McCLELLAND, J.L. & KAWAMOTO, A.H. (1986). Mechanisms of sentence processing: Assigning roles to constituents of sentences. In J.L. McClelland, D.E. Rumelhart and the PDP Research Group (Eds.) Parallel Distributed Processing:  Explorations in the Microstructure of Cognition. Vol 2. Cambridge MA: The MIT Press. NEUFELDT, V. (Ed.). (1988). Webster's New World Dictionary of  American English. Third College Edition. New York: Webster's New World. NEWPORT, E., GLEITMAN, H., & GLEITMAN, E. (1977). Mother, I'd rather do it myself: Some effects and noneffects of maternal speech style. In C.E. Snow & C.A. Ferguson, (Eds.) Talking  to Children: Language Input and Acquisition. New York: Cambridge University Press.  Metaphor Comprehension - 153 NIMS, M.F. (Trans.). (1967). Poetria Nova of Geoffrey of Vinsauf. Toronto: Pontifical Institute of Mediaeval Studies. ORTONY, A. (1979). Metaphor: A multidimensional problem. In A. Ortony (Ed.) Metaphor and Thought. Cambridge: Cambridge University Press. ORTONY, A., SCHALLERT, D.L., REYNOLDS, R.E., & ANTOS, S.J. (1978). Interpreting metaphors and idioms: Some effects of context on comprehension. Journal of Verbal Learning and Verbal Behavior, 17, 465-477. POLLIO, H.R., FABRIZI, M.S., SILLS, A., & SMITH, M.K. (1984). Need metaphoric comprehension take longer than literal comprehension? Psycholinguistic Research, 13, 195-214. POSNER, M. (1978). Chronometric Explorations of Mind. Hillsdale NJ: Erlbaum. RICHARDS, I.A. (1936). Metaphor. In I.A. Richards The Philosophy of Rhetoric. London: Oxford University Press. ROSCH, E. (1973). On the internal structure of perceptual and semantic categories in cognitive development. In T.E. Moore (Ed.) Cognitive Development and the Acquisition of Language. New York: Academic Press. ROSCH, E. (1978). Principles of categorization. In E. Rosch & B.B. Lloyd, (Eds.) Cognition and Categorization. Hillsdale, NJ: Erlbaum. RUBENSTEIN, H., GARFIELD, L., & MILLIKAN, J.A. (1970). Homographic entries in the internal lexicon. Journal of Verbal Learning- and- Verbal Behavior. S, 487-49-4.  Metaphor Comprehension - 154 RUMELHART, D.E. & ORTONY, A. (1977). The Representation of Knowledge in Memory. In R.C. Anderson, R.J. Spiro, & W.E. Montague (Eds.) Schooling and the Acquisition of Knowledge. Hillsdale, N.J.: Lawrence Erlbaum Associates. SCHANK, R. (1982). Dynamic Memory: A Theory of Learning in Computers and People. New York: Cambridge University Press. SCHANK, R. (1990). Tell me a Story: A New Look at Real and Artificial Memory. New York: Charles Scribner and Sons. SEARLE, J.R. (1979). Metaphor. In A. Ortony (Ed.) Metaphor and Thought. Cambridge: Cambridge University Press. SWINNEY, D.A. (1979). Lexical access during sentence comprehension: (Re)consideration of context effects. Journal of Verbal Learning and Verbal Behavior, 18, 645-659. SWINNEY, D.A. & CUTLER, A. (1979). The accessing and processing of idiomatic expressions. Journal of Verbal Learning and Verbal Behavior, 18, 523-534. TABOSSI, P. (1988). Accessing lexical ambiguity in different types of sentential contexts. Journal of Memory and Language, 27, 324-340. TARTTER, V.C. (1986). Language Processes. New York: Holt, Rinehart, & Winston. TVERSKY, A. (1977). Features of Similarity. Psychological Review, 84, 327-52.  Metaphor Comprehension - 155 VERBRUGGE, R.R., & McCARRELL, N.S. (1977). Metaphoric comprehension: Studies in reminding and resembling. Cognitive Psychology, 9, 494-533. WHORF, B.L. (1956). An American Indian model of the universe. In J.B. Carroll (Ed.) Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. New York: Wiley. WITTGENSTEIN, L. (1953). Philosophical Investigations. (Trans.) G.E.M. Anscombe. Oxford: Blackwell.  Metaphor Comprehension - 156  APPENDICES  Metaphor Comprehension - 157  APPENDIX A  158  SENTENCE COMPREHENSION Odie Geiger - Psychfest, 1990 March, 1990 This experiment involves the reading and understanding of short groups of sentences. On each trial, a few sentences will be displayed, followed by a word or phrase at the end to complete the meaning. Your task will be to read the sentences carefully, and when you are sure that you understand their meaning up to that point, push the spacebar at the bottom of the keyboard. Immediately, the sentences will disappear and the ending will be displayed. This portion of the experiment will be timed. As soon as you understand the ending so that the whole presentation makes sense, push the spacebar again. A new group of sentences will then appear on the screen and the procedure will continue through 160 trials. It is important to the success of the experiment that you understand the meaning of the entire presentation. In order to ensure that you are reading the sentences carefully enough for your responses to be meaningful, a short quiz will follow the experiment, to find out if the meanings were clear to you. This does not mean that the sentences will be tricky or difficult. You will likely understand the meaning of the sentences as soon as you read them. Nor does it mean that you need to memorize the sentences, the details, or the names of people who may be incidentally mentioned. It is just a way of keeping you from responding automatically as soon as the final phrase is flashed on the screen. Since the initial portion of each presentation is not timed, you can leave it on the screen for as long as you like to make sure that the meaning is clear to you. After you press the spacebar, you should respond as quickly as possible, once the meaning of the whole is clear. We will provide a few practice trials before the experiment begins to accustom you to the response procedure. You will have eight sets of 20 trials each to complete. If you become tired at any point, please feel free to request a break between sets. You will be given a point of class credit and an explanation of what the experiment is about after your sessions have been completed. Thank you for your interest and participation. Odie Geiger Juliet Armstrong Research Assistants  159  IDIOMS1.APX^ EXPERIMENT 1 - IDIOMS  1)II context: Dean spoiled the surprise that Joan had been planning for their mother's birthday. When he realized what he had done, he apologized for having 1)II target: let the cat out of the bag. 1)IP target: revealed the secret. 1)IS target: fallen asleep in the closet. 1)IL context: Walking back from the store, Anne found a kitten which she put in with her groceries. She got home and her puppy went wild when she 1)IL target: let the cat out of the bag. 2)11 context: Watching the young men play football in the park on an early spring evening made Ralph nostalgic. He found himself wishing that he could 2)11 target: turn back the clock. 2)IP target: return to his youth. 2)IS target: spend spring somewhere else. 2)IL context: When he arrived an hour late for work Monday morning, Barry remembered that Daylight Savings Time had begun the night before. As soon as he got home, he went into his bedroom to 2)IL target: turn back the clock. 3)11 context: Mary found it difficult to make friends at her new school. Finally, she decided to hold a party as a way to 3)11 target: break the ice. 3)IP target: become acquainted. 3)IS target: try some new recipes. 3)IL context: After running his team of horses for six hours over rough frozen roads, Pierre arrived at an old barn with a watering trough outside. His next task was to find a stick heavy enough to 3)IL target: break the ice. 4)11 context:George and Martha had a good relationship, because they recognized that neither was perfect. They both knew that Martha had a tendency to nag, and that George would occasionally 4)11 target: fly off the handle. 4)IP target: lose his temper. 4)IS target: read yachting magazines. 4)IL context: Benny was a tame parrot who was often allowed to fly around the house. One of his favourite perches was right above the kitchen sink. In order to turn on the water, it was first necessary to make Benny 4)IL target: fly off the handle.  IDIOMS1.APX^  160  5)11 context: Curtis was an irresponsible gambler but he was saved by hi charming manner, and an incredible amount of good luck. No matter how much difficulty he got himself into, he always seemed able to 5)11 target: land on his feet. 5)IP target: recover from his troubles. 5)IP target: write scholarly essays. 5)IL context: Mark had studied gymnastics since he was six. He was the most graceful athlete at his club, and the most talented. His routines were both spectacular and controlled, and the audiences knew that he would always 5)IL target: land on his feet. 6)11 context: When Mr.Wilson retired, he put his money into savings bonds An investment counselor had suggested a way to earn more interest, but Mr. Wilson thought that he 6)11 target: could smell a rat. 6)IP target: might be cheated. 6)IS target: should help the government. 6)IL context: Linda hated the big old house she inherited. Every time sh4 walked down the dirty, dingy hallway she heard noises and she was sure that she 6)IL target: could smell a rat. 7)11 context: The contact lens popped out somewhere on the basketball court. The referee had to stop the game while the players went over the court 7)11 target: with a fine-toothed comb. 7)IP target: carefully and completely. 7)IS target: with a vacuum cleaner. 7)IL context: After swimming, Margaret's long hair was a tangled mess. She always had to spend extra time drying it, brushing it, and finally, combing it 7)IL target: with a fine-t000thed comb. 8)11 context: The wedding was scheduled to take place in the garden that afternoon. By 10 a.m., dark clouds began gathering. All the family could do for the next four hours was to 8)11 target: hold their breath. 8)IP target: wait helplessly. 8)IS target: finish remodelling the garage. 8)IL context: One of the techniques Carmen had learned as a lab technician was that patients would become less nervous during an injection if they were told to look away and 8)IL target: hold their breath.  IDIOMS1.APX^  161  9)11 context: When her boss walked in her door with a big smile on his face, Leslie knew she had been promoted. Weeks earlier, rumours about management changes had been 9)11 target: in the wind. 9)IP target: topics of speculation. 9)IS target: about planned layoffs. 9)IL context: The sky grew grey and the temperature dropped sharply. Huc raindrops splashed against the pavement, and piles of leaves swirlec about 9)IL target: in the wind. 10)11 context: When John finally decided to do something about his lack c physical fitness, he went to visit his doctor, who recommended a program that was 10)II target:tailor-made for him. 10)1P target:exactly appropriate for him. 10)IS targt:designed for pregnant women. 10)IL context: Oscar was tall and extremely thin, but despite this he always looked elegant and powerful in his suits because they were 10)IL target:tailor-made for him. 11)II context: All of Ben's earnings went toward financing his annual holiday in Reno, where he consistently 11)II target: lost his shirt. 11)IP target: gambled and lost. 11)IS target: gave money to losers. 11)IL context: The last time Ricky went swimming in the pond, a wind came up and scattered his clothes. He finally found his pants, but he had to confess to his mother that he had 11)IL target: lost his shirt. 12)11 context: Bob's parents and teachers had never had any trouble with him as a child, but when he became a teenager he suddenly became 12)11 target: too big for his boots. 12)IP target: arrogant and disrespectful. 12)IS target: expert at needlepoint. 12)IL context: Carl's parents were shocked at how much it was costing to feed and clothe their growing son. They bought him a new outfit for school, but within only three months, he had become 12)IL target: too big for his boots. 13)11 context: The ad offered a 1968 Corvette in excellent condition for $2000. When the owner agreed to an offer over the phone of $1500, Cynthia hurried over to his house to 13)11 target: nail it down. 13)IP target: confirm the deal. 13)IS target: see what color it was. 13)IL context: The tornado caused considerable damage, but the only damage to the Johnson's house was a single loose shingle. -Kr Johnson had to clinib up to the roof to 13)IL target: nail it down.  IDIOMS1.APX  ^  162  14)11 context: Charles disliked working with Neil, who wouldn't do his share and yet attempted to take credit for the completed project. Finally, Charles decided that it was 14)11 target: a game two could play. 14)IP target: a strategy he would try. 14)IS target: wrong to dislike Neil. 14)IL context: Jean and Sheila lived on a farm far away from children their own age. When they received a new video game at Christmas, they were excited to find that it included 14)IL target: a game two could play. 15)11 context: Everybody had a good time at the Spring Dance; the fact that the caterers got lost and the food didn't arrive until 11 p.m. was the only 15)11 target: fly in the ointment. 15)IP target: disruptive event. 15)IS target: reason Mary cried. 15)IL context: Paula's opinion about the dubious joys of camping was confirmed after she got severely burned one day. When she went to find something to relieve her agony, she discovered a 15)IL target: fly in the ointment. 16)11 context: The plant had suffered a number of financial losses as world markets had changed. The employees didn't want to start looking for other jobs, but they could 16)11 target: see the writing on the wall. 16)IP target: predict the consequences. 16)IS target: organize a co-op. 16)IL context: The kids had a great time scribbling while their babysitter was talking on the telephone. When they finished they went outside, hoping that she wouldn't 16)IL target: see the writing on the wall. 17)11 context: All of Margaret's friends warned her against marrying Brian, but none were sympathetic when the marriage failed. They all felt that Margaret had entered the marriage 17)11 target: with her eyes wide open. 17)IP target: knowing the likely consequences. 17)IS target: for money and revenge. 17)IL context: Tammy's friends wanted to see her even before she had fully recovered from the anesthetic, but the doctor suggested that she would be more comfortable if she could visit them 17)IL target: with her eyes wide open. 18)11 context: Max had borrowed money to buy a house, a car, and a boat before he lost his job. Almost instantly, he discovered that his creditors were 18)11 target: breathing down his neck. 18)IP target: demanding repayment. 18)IS target: generous and sympathetic. 18111, context: Adam enjoyed having a cat in the house. It slept in the bed with him, and he found it comforting to fall asleep while listening to it purr and feeling it 18)IL target: breathing down his neck.  IDIOMS1.APX^  163  19)11 context: Harold loved his new job and was glad to be rid of the pressures of his old one. However, he knew that with the salary c he had taken, he would have to 19)11 target: tighten his belt. 19)IP target: cut his expenses. 19)IS target: work extra hours. 19)IL context: Paul was pleased with having lost 15 pounds. At first, h thought he might have to buy new clothes, but he discovered that h old ones would be fine if he would just 19)IL target: tighten his belt. 20)11 context: Chris and his friends were fooling around when they accidentally broke his mother's treasured crystal vase. Several days passed before his mother noticed, and by then Chris was relieved to 20)11 target: get it off his chest. 20)IP target: get his cuts treated. 20)IL context: Elmer was working on his car when the jack slipped, pinning him underneath. He was not seriously injured, but it took three rescue workers to 20)IL target: get it off his chest.  METAPHOR.APX  ^  164  EXPERIMENT 1 - METAPHORS 1)MM context: The committee decided to wait six months longer before evaluating Laura's performance because they felt she was still 1)MM target: green. 1)MP target: inexperienced. 1)MS target: recovering. 1)ML context: Joanne thought that Ted would notice that she had painted their bedroom a different color, but to him it was still 1)ML target: green. 2)MM context: A large proportion of Canadians feel that the Free Trade Agreement will mean the end of 2)MM target: the Maple Leaf. 2)MP target: Canada. 2)MS target: Nanaimo Bars. 2)ML context: One of my favourite memories of childhood is of sitting out under the trees in early fall, and marvelling at the range of colour and texture in 2)ML target: the maple leaf. 3)MM context: Corrine had three classes that morning, with a quiz in one and a presentation in another. By noon she felt that she definitely needed to stop for 3)MM target: fuel. 3)MP target: food. 3)MS target: exercise. 3)ML context: Art enjoyed the mental challenge of car rallies even more than the physical. One of his greatest delights was to keep a constant record of his car's consumption of 3)ML target: fuel. 4)MM context: Angela's parents disliked her smoking. They warned her that besides being unattractive, cigarette smoking could be a 4)MM target: time bomb. 4)MP target: latent killer. 4)MS target: growth inhibitor. 4)ML context: When a child refused to pass his teddy bear through the metal detector at the airport, the official forcibly took the bear away. He explained that any object had to be considered a potential 4)ML target: time bomb. 5)MM context: The politician's presentation was smooth, but the facts he presented were not consistent with what Terry knew to be true. She suspected that the data had been 5)MM target: cooked. 5)MP target: changed. 5)MS target: stolen. 5)ML context: Marilyn spent the morning in the orchard, picking apples for applesauce. She didn't step working until evening, when the apples were finally peeled, cored, and 5)ML target: cooked. -  —  METAPHOR.APX^  165  6)MM context: It is important to realize that any crative endeavor necessarily produces frustrations and mistakes. Not even Einstein ideas were all 6)MM target: gems. 6)MP target: valuable. 6)MS target: his own. 6)ML context: Buying jewelry today is tricky. With so many synthetics available, the unwary buyer can easily be misled into buying artificial materials, believing that they are 6)ML target: gems. 7)MM context: Bert warned his new neighbors to stay away from Dr. McNeil. He had been sued twice for malpractice, and had the reputation in town of being 7)MM target: a butcher. 7)MP target: incompetent. 7)MS target: a peeping tom. 7)ML context: The Hills were delighted to find a tiny shopping district near their new house which featured a produce market, a coffee shop a baker, and 7)ML target: a butcher. 8)MM context: It was a perfect summer evening. The lake was deserted but we could hear the loons calling. The last rays of the sun made the lake a shimmering 8)MM target: sheet of foil. 8)MP target: sheet of light. 8)MS target: mass of fish. 8)ML context: Cleaning up after the party was as big a job as preparing for it. Owen vacuumed the living room, while Liz wrapped up the leftovers in a 8)ML target: sheet of foil. 9)MM context: The babysitter grew increasingly annoyed as the crashing sounds and yelling increased. Finally she decided it was time to discipline the 9)MM target: troops. 9)MP target: children. 9)MS target: dogs. 9)ML context: The soldiers had been in the jungle for six months and were suffering from disease and fatigue. The general decided it was time to replace them with fresh 9)ML target: troops. 10)MM context: Fluffy was older than Brenda's children, and they regarded her as a large, dignified cat, while Brenda could still remember her as a tiny striped 10)MM target: ball. 10)MP target: kitten. 10)MS target: tiger. ___111)MT,centext: Tire game was over almost before it had started because Ernie hit a grand slam home run, and afterwards nobody could find the 10)ML target: ball.  METAPHOR.APX  ^  166  11)MM context: Karen came into the office wearing a new outfit almost every week. She expected us to compliment her, but everyone got tired of her continual 11)MM target: fishing. 11)MP target: hinting. 11)MS target: role-playing. 11)ML context: The highlight of my summers was always Uncle Barney's visit, when he would take my brothers and me up north for a week ti do some 11)ML target: fishing. 12)MS context: After Kenley lost his job, he decided that he was just as happy living on welfare. His old friends knew that his indifferenc was only a 12)MM target: facade. 12)MP target: pretense. 12)MS target: symptom. 12)ML context: The architects decided that the building would have to be rebuilt, but that they could construct a whole new building and maintain the old 12)ML target: facade. 13)MM context: When Peter married Ellen, it was because she had enough money to provide him with whatever he wanted. Ellen had no idea that he was so 13)MM target: calculating. 13)MP target: sly. 13)MS target: clever. 13)ML context: James always dreaded his monthly bookkeeping. He was never good at math, and his balances often came out wrong even after hours of 13)ML target: calculating. 14)MM context: The police were reluctant to call off their search for the missing child until they were sure that the area had been completely 14)MM target: scoured. 14)MP target: searched. 14)MS target: quarantined. 14)ML context: Tina was the best housekeeper we ever had. She took pride in her work, and liked to see floors waxed, silver polished, and sinks 14)ML target: scoured. 15)MM context: It was bad enough to lose $135, but Chuck also had to replace all his credit cards after he discovered his wallet had been 15)MM target: ripped off. 15)MP target: stolen. 15)MS target: burned. 15)ML context: Actors often wear clothes that are held together with Velcro for those moments when the script calls for their costumes to be 15)ML target: ripped off. -  -  METAPHOR.APX^  167  16)MM context: It was time to call a strike. None of the changes that management proposed had been made, and the workers were starting i make 16)MM target: noises. 16)MP target: accusations. 16)MS target: mistakes. 16)ML context: The mechanics at Mike's Garage amused one another by imitating the sounds customers would make as they tried to describ their cars's odd 16)ML target: noises. 17)MM context: After the disasterous 1929 stock market crash, a lot of formerly prosperous financiers saw their future prospects suddenly turn 17)MM target: black. 17)MP target: bleak. 17)MS target: illegal. 17)ML context: Fashion is fickle. At different times, both sophisticated society and rebellious youth have focused their tastes around the color 17)ML target: black. 18)MM context: The studio audiences seemed to love the new game show host, but the critics and a number of the people who watched the show felt that he was too 18)MM context: plastic. 18)MP target: artificial. 18)MS target: short. 18)ML context: One of the biggest problems faced by current environmentally conscious governments is what can be done to stop the production of throwaway 18)ML target: plastic. 19)MM context: Alice's playmates thought it was funny to pull the beetle's legs off one by one and watch it struggle to walk, but Alice thought it was 19)MM target: sick. 19)MP target: disgusting. 19)MS target: inspiring. 19)ML context: It is often a difficult challenge for working couples to rearrange their day in order to get adequate home care when their child is 19)ML target: sick. 20)MM context: There has been a lot of inefficiency in our administration of funds. Our new policy is to manage all our funding under a single 20)MM target: umbrella. 20)MP target: agency. 20)MS target: dictator. contemn It was a grey, rainy day and Harriet dozed on the bus until she reached her stop. After she had gotten off, she realized that she had left behind her 20)ML target: umbrella.  ^ ^  ^ SENTENCE COMPREHENSION Odie Geiger - Psychfest, 1990  168  name: ^ major: ^ class credit to:(instructor) ^ Is^English your most familiar language? ^ Please answer the following questions with "T" if the meaning on the question is the same as one of the items you were given to read, or "F" if its meaning is different. Some of the items in the experiment were presented more than once with different endings; as long as the question is true of ONE of those you remember, respond "T". Once again, there are no trick questions here. For example, names are not switched between items, and incidents have not been combined. 1. ^ Actors wear clothing that is made to come apart easily. 2. ^ The kids built a miniature farm while their babysitter was on the telephone. 3. ^ Alice's playmates tried to get beetles to fight over bits of grain. ^4. ^ The house Linda inherited smelled of rats. 5. ^ Mary joined the school newspaper as a way of making new friends. 6. ^ The housebound wedding party ended up remodelling the garage. 7. ^ In the winter sun, the lake looked like a white tablecloth. 8. ^ The Free Trade agreement could mean the end of Nanaimo Bars. 9. ^ The politicians presentation was smooth but his facts seemed suspicious. 10. ^ Barry was late for work because he got a traffic ticket for speeding. 11. ^ Charles decided to use the same strategy as Neil, and leave the work to someone else. 12. ^ Max bought his house, car,*and boat with money he inherited. 13. ^ George and Martha both had faults, but they had a good relationship in spite of them. 14. ^ Kenley's friends felt that his contentment with welfare was a symptom of his distress. 15. ^ Brenda's children loved watching the cat chase a red ball of yarn down the stairs. 16. ^ The new game show host impressed the critics as highly talented. 17. ^ It was felt that funds could be better administered under the direction of a single agency. 18. ^ When John asked about a program of physical fitness, the doctor said he was in fine shape. 19. ^ Dean apologized to Joan for revealing the se-ret of her plans for a surprise party. 20. ^ Black has been the color of dress for both the sophisticated and the rebellious. 21. ^ The mechanics at Mike's garage amused themselves with video movies.  ^  169  ^22. ^ Her office mates felt that Karen was fishing for compliments on her new outfits. 23. ^ Carmen told patients who were having laboratory work done to look away and hold their breath. 24. ^ Ralph joined the young men playing football in the park. 25. ^ The police refused to continue their search for the missing child. 26. ^ Ben regularly lost his money while gambling in Reno. 27. ^ When the child refused to pas his teddy bear through the metal detector, the officers laughed. 28. ^ When Ellen married Peter, she had no idea that he was interested in her money. 29. ^ Barry became arrogant and disrespectful when he became a teenager. 30. ^ When Ricky went swimming, he was bitten all over by minnows. 31. ^ All Einstein's ideas were spectacularly successful. 32. ^ Corrine was hungry after a busy morning of classes.  Metaphor Comprehension - 170  APPENDIX B  SENTENCE COMPREHENSION 3 Odie Geiger September, 1990  ^171  This experiment involves the reading and understanding of short groups of sentences. On each trial, a few sentences will be displayed, followed by a word or two to complete their meaning. Your task will be to read the sentences carefully, and when you are sure that you understand their meaning up to that point, push one of the buttons on the response box. Immediately, the sentences will disappear and the ending will be displayed. This portion of the experiment will be timed. As soon as you understand the ending so that the whole presentation makes sense, push the button again. A new set of sentences will then appear on the screen and the procedure will continue through 160 trials. It is important to the success of the experiment that you understand the meaning of the entire presentation. In order to ensure that you are reading the sentences carefully enough for your responses to be meaningful, a short quiz will follow the experiment. This does not mean that the sentences will be tricky or difficult. You will likely understand them as soon as you read them. Nor does it mean that you need to memorize the sentences, the details, or the names of the people who may be incidentally mentioned. It is just a way of keeping you from responding automatically as soon as the final words appear on the screen. Since the initial portion of the presentation is not timed, you can leave it on the screen as long as you like to make sure that its meaning is clear to you. After you press the button to display the ending, you should respond as quickly as possible with a second press of the button when the meaning of the whole phrase is understood. We will provide a few practice trials before the experiment begins to accustom you to the response procedure. You will have eight sets of 20 trials each to complete. If you become tired at any point, please feel free to take a break between sets. You will be given a point of class credit and an explanation of what the experiment is about when the session is over. Thank you for your interest and participation. Odie Geiger Graduate Student  DEADMET2.APX^  172  EXPERIMENT 2 - DEAD METAPHORS 1)DD context: The mountaineers decided that they needed two camps to stoi supplies for their climb. One would be midway up the mountain, and the other would be at the 1)DD target: foot. 1)DP target: base. 1)DS target: castle. 1)DL context: A whole day of Christmas shopping was too much for Jessica. When she got home she discovered that she had developed a huge blister on her 1)DL target: foot. 2)DD context: Tim's teachers were impressed with him from the start. He h, the very appealing advantage of being both very personable and very 2)DD target: bright. 2)DP target: smart. 2)DS target: wealthy. 2)DL context: After nearly two weeks of rain and overcast skies, Myrtle wi shocked to look outside her window and see that the sky was clear anc 2)DL target: bright. 3)DD context: Steve tried to establish a good relationship with his new mother-in-law, but despite all his efforts, she remained 3)DD target: cold. 3)DP target: unfriendly. 3)DS target: shy. 3)DL context: Alana loved the idea of going to study as an exchange studen in Norway. All she knew about it was that it would be 3)DL target: cold. 4)DD context: The camera crew worked hard continuously from 6 a.m. til noon, when the director called for a much needed 4)DD target: break. 4)DP target: rest. 4)DS target: swim. 4)DL context: Kenny didn't know the road was under construction until his truck bounced into the air, causing most of the 200 dozen eggs he was carrying to 4)DL target: break. 5)DD context: All Roger could talk about was cars and baseball. Sally felt that she had never before met anyone so 5)DD target: shallow. 5)DP target: uninformed. 5)DS target: masculine. 5)DL context: The beach at Lost Lake is perfect for young children. It extends for meters and yet remains sandy, clear and 5)DL target: shallow.  DEADMET2.APX^  173  6)DD context: After Kenley lost his job, he decided that he was just as happy living on welfare. His old friends knew that his indifference was only a 6)DD target: facade. 6)DP target: pretense. 6)DS terget: symptom. 6)DL context: The architects decided that the building would have to be rebuilt, but that the could construct a whole new building and maintain the lovely old 6)DL target: facade. 7)DD context: Whenever my problems seem to big to handle, I always know that I can turn to my Uncle Dave for 7)DD target: support. 7)DP target: guidance. 7)DS target: laughs. 7)DL context: One of the most important considerations for engineers is making sure that superstructures have adequate 7)DL target: support. 8)DD context: Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were 8)DD target: firm. 8)DP target: definite. 8)DS target: progressive. 8)DL context: Cheeses have been developed with flavours varying from mild to tangy, and textures ranging from soft to 8)DL target: firm. 9)DD context: Jason was a superb interior decorator. It was easy to recognize rooms which had his distinctive 9)DD target: touch. 9)DP target: style. 9)DS target: clutter. 9)DL context: Good doctors soon develop a sense of how to diagnose their patients' problems. They know just what questions to ask, and which areas to 9)DL target: touch. 10)DD context: The United States represents a society in which the cultures of a number of different nations have become 10)DD target: interwoven. 10)DP target: combined. 10)DS target: debased. 10)DL context: While travelling through India, Lois bought several meters of beautiful silk fabrics into which strands of metallic threads had been 10)DL target: interwoven.  DEADMET2.APX^  174  11)DD context: Whenever Toby gets a new mystery novel by her favourite  author, she puts everything else aside because she finds it so 11)DD target: absorbing. 11)DP target: interesting. 11)DS target: difficult. 11)DL context: The major oil companies are interested in developing a material for cleaning up oil spills that is harmless, yet highly o 11)DL target: absorbing. 12)DD context: Many experienced cyclists are terrified of riding in Soutl  America, where roads are in poor repair and often have inadequate target: shoulders. target: edges. target: restrooms. context: As Grandma got older, she began to wear more warm clothes, even in the summertime. She never went out without putting a shawl over her 12)DL target: shoulders. 12)DD 12)DP 12)DS 12)DL  13)DD context: When Pam told Helen that she had eloped with Max against  her parent's wishes and that they were now married, Helen was target: shocked. target: surprised. target: jealous. context: It is almost certain that if you touch an electrical appliance while you are in the bathtub you will get 13)DL target: shocked. 13)DD 13)DP 13)DS 13)DL  14)DD context: In an antique shop in the Maritimes, Jefferey discovered a  beautiful old clock with a handpainted target: face. target: dial. target: barometer. context: Emily could scarcely believe that an eight year old child would continually try to sneak off to bed without washing his hands and 14)DL target: face. 14)DD 14)DP 14)DS 14)DL  15)DD context: Rex liked the idea of living in New York but he worried  about his health because the smell of industrial pollution was so target: strong. target: intense. target: disguised. context: Casey chose his food carefully, exercised regularly, and took vitamin supplements in order to keep himself healthy and 15)DL target: strong. 15)DD 15)DP 15)DS 15)DL  DEADMET2.APX^  175  16)DD context: Although she did not like to admit it, Bianca felt that ii  many ways her life and her mother's life were target: parallel. target: alike. target: cursed. context: It must have been very difficult for the workers who built the original Canadian National Railway to keep the tracks exactly 16)DL target: parallel. 16)DD 16)DP 16)DS 16)DL  17)DD context: The lawyers had assembled a highly detailed and carefully  documented case against the suspected drug dealers. The judge felt the facts were 17)DD target: clear. 17)DP target: correct. 17)DS target: distorted. 17)DL context: Bermuda is a great place to scuba dive. There is a huge variety of exotic tropical fish, and the water is warm and 17)DL target: clear.  18)DD context: Our neighbors called the police when the 30 or 40 teenagers  attending the party at the end of the block suddenly got target: wild. target: rowdy. target: quiet. context: It is interesting for people to be able to see animals in zoos, but it is not the same experience as seeing the animals when they are 18)DL target: wild. 18)DD 18)DP 18)DS 18)DL  19)DD context: Nigel couldn't understand how Professor Rogers could get so  much attention for his theory when his reasoning seemed so target: weak. target: unconvincing. target: ethnocentric. context: The hostages were released after four years in captivity. They showed the effects of inadequate food and excercise, and were extremely 19)DL target: weak. 19)DD 19)DP 19)DS 19)DL  20)DD context: When Peter married Ellen, it was because she had enough  money to provide him with whatever he wanted. Ellen had no idea that he was so 20)DD target: calculating. 20)DP target: ruthless. 20)DS target: clever. 20)DL context: James always dreaded his monthly bookkeeping. He was never very good at math, and his balances often came out wrong even after hours of 20)DL target: calculating.  MODMET2.APX^  176  EXPERIMENT 2 - METAPHORS 1)MM context: The committee decided to wait six months longer before evaluating Laura's performance because they felt she was still 1)MM target: green. 1)MP target: inexperienced. 1)MS target: recovering. 1)ML context: Joanne thought that Ted would notice that she had painted their bedroom a different color, but to him it was still 1)ML target: green. 2)MM context: As soon as our garage sale opened, a tall young man with a lot of money appeared. He looked over everything carefully, and bought up all the 2)MM target: cream. 2)MP target: valuables. 2)MS target: pillows. 2)ML context: Some people like milk in their tea and some people like lemon, or honey, or sugar, but Marilyn is the only person I know who drinks her tea with 2)ML target: cream. 3)MM context: Corrine had three classes that morning, with a quiz in one  and a presentation in another. By noon, she felt that she definitely needed to stop for 3)MM target: fuel. 3)MP target: food. 3)MS target: exercise. 3)ML context: Art enjoyed the mental challenge of car rallies even more than the physical. One of his greatest delights was to keep a constan record of his car's consumption of 3)ML target: fuel. 4)MM context: When Joel turned 22 his parents insisted that he find an apartment and get a job. They were afraid that he would spend the rest of his life as a 4)MM target: parasite. 4)MP target: dependent. 4)MS target: hermit. 4)ML context: Albee's greatest fear about going to Africa to work with starving people was that he might be exposed to some deadly intestinal 4)ML target: parasite. 5)MM context: The politician's presentation was smooth, but the facts he  presented were not consistent with what Terry knew to be true. She suspected that the data had been 5)MM target: cooked. 5)MP target: changed. 5)MS target: stolen. 5)ML context: Marilyn spent the morning in the orchard, pinking apples for applesawee. She didnTt stop working until evening, when the apples were finally peeled, cored, and 5)ML target: cooked.  MODMET2.APX^  177  6)MM context: It is important to realize that any creative endeavour necessarily produces frustrations and mistakes. Not even Einstein's ideas were all 6)MM target: gems. 6)MP target: valuable. 6)MS target: original. 6)ML context: Buying jewelry today is tricky. With so many synthetics available, the unwary buyer can easily be mislead into buying artificial materials, believing that they are 6)ML target: gems. 7)MM context: Christy couldn't wait to get to the beaches in Hawaii. She bought a new bikini, but she was determined to be careful. She knew how easily she could 7)MM target: toast. 7)MP target: sunburn. 7)MS target: drown. 7)ML context: It is really hard to adjust to a full English breakfast of eggs, bacon, sausage, kippers, tomatoes, and mushrooms, if your usua morning meal consists of coffee and 7)ML target: toast. 8)MM context: As the newest teacher on staff, Rebecca willingly accepted responsibility for the school's Christmas pageant. By October, the project was becoming a 8)MM target: bear. 8)MP target: burden. 8)MS target: farce. 8)ML context: When my cousins from New York came to visit, the thing they wanted to do most was to camp out in the Rockies so they could photograph a live 8)ML target: bear. 9)MM context: The babysitter grew increasingly annoyed as the crashing sounds and yelling increased. Finally she decided it was time to discipline the 9)MM target: troops. 9)MP target: children. 9)MS target: dogs. 9)ML context: The soldiers had been in the jungle for six months and were suffering from disease and fatigue. The general decided it was time to replace them with fresh 9)ML target: troops. 10)MM context: Don's chemistry class was asked to attend a series of lectures by a Nobel Prize winning French chemist. Don went, but found the lectures difficult to 10)MM target: unpack. 10)MP target: understand. 10)MS target: translate. context: It was incredible that a full year after they had moved, the Gordon's had boxes in their basement that they still hadn't taken the time to 10)ML target: unpack.  MODMET2.APX^  178  11)MM context: Karen came into the office wearing a new outfit almost ev week. She expected us to compliment her, but everyone got tired of her continual 11)MM target: fishing. 11)MP target: hinting. 11)MS target: theatrics. 11)ML context: The highlight of my summers was always Uncle Barney's vis. when he would take my brothers and me up north for a week to do sox 11)ML target: fishing. 12)MM context: Andrea tried to enhance her lectures with trips and  demonstrations, worked with parents, and coached volleyball. By the time she got home each night, she was 12)MM target: empty. 12)MP target: fatigued. 12)MS target: sweaty. 12)ML context: Paul hoped he would have a chance to talk to one of the speakers after the conference. He hurried back to the auditorium, but he found it 12)ML target: empty. 13)MM context: Ross thought he had a good chance of being the next club  president, but from the standpoint of the executive committee, he wasn't even in the 13)MM target: picture. 13)MP target: candidates. 13)MS target: club. 13)ML context: My neighbor took up photography after he retired. He is extremely talented, and many people have already asked him if they could buy a 13)ML target: picture. 14)MM context: Carmen had always tried to think for herself. After she got  too drunk at an office party, she wondered how she had been able to follow the 14)MM target: wave. 14)MP target: crowd. 14)MS target: music. 14)ML context: The dangers of an earthquake are even greater for people living along a coastline who also have to worry about the possibilit] of a tidal 14)ML target: wave. 15)MM context: Jason had produced a demonstration recording when he was a teenager. One day he heard a new group on MTV and realized that his old lyrics had been 15)MM target: ripped off. 15)MP target: stolen. 15)MS target: drivel. 15)ML context: Actors often wear clothes that are held together with Velcro for those moments when the script calls for their costumes to be 15)ML target: ripped off.  MODMET2.APX  ^  179  16)MM context: It was time to call a strike. None of the changes that management proposed had been made, and the workers were starting ti make 16)MM target: noises. 16)MP target: accusations. 16)MS target: mistakes. 16)ML context: The mechanics at Mike's Garage amused one another by imitating the sounds customers would make as they tried to describc their car's odd 16)ML targets: noises. 17)MM context: After the disastrous 1929 stock market formerly prosperous financiers saw their future turn 17)MM target: black. 17)MP target: bleak. 17)MS target: illegal. 17)ML context: Fashion is fickle. At different times, society and rebellious youth have focused their color 17)ML target: black.  crash, a lot of prospects suddenly  both sophisticated tastes around the  18)MM context: The studio audiences seemed to love the new game show host, but the critics and a number of people who watched the show felt thz he was too 18)MM target: plastic. 18)MP target: artificial. 18)MS target: short. 18)ML context: One of the biggest problems faced by current environmentall conscious governments is what can be done to stop the production of throwaway 18)ML target: plastic. 19)MM context: Alice's playmates thought it was funny to pull the beetle's legs off one by one and watch it struggle to walk, but Alice thought it was 19)MM target: sick. 19)MP target: disgusting. 19)MS target: inspiring. 19)ML context: It is often a difficult challenge for working couples to rearrange their day in order to get adequate home care when their child is 19)ML target: sick. 20)MM context: There has been a lot of inefficiency in our administration of funds. Our new policy is to manage all of our funding under a single 20)MM target: umbrella. 20)MP target: agency. 20)MS target: bureaucrat. 20)ML context: It was a grey, rainy day and Harriet dozed on the bus until she reached her stop. After she had gotten off, she realized that she had left behind her target: umbrella. 20)ML  ^  ^ EXPERIMENT 2 SENTENCE COMPREHENSION Odie Geiger  180  name: ^ major: ^ class credit to (instructor): ^ Is^English your most familiar language? ^ Please answer the following questions with a "T" if the meaning of the question is the same as that of one of the items you were given to read, or "F" if the meaning is different. Some of the items in the experiment were presented more than once with different endings; as long as the question is true of any ONE of those you remember, respond "T". Once again, there are no trick questions here. For example, names are not switched between items, and incidents have not been combined or altered. 1. ^ Actors wear clothing that is made to come apart easily. 2. ^ Steve wanted to have a good relationship with his new mother-inlaw. 3. ^ Alice's playmates tried to get the beetles to fight over bits of grain. 4. ^ All Alana knew about Norway was that it was a cold country. 5. ^ Rebecca found that producing the school's Christmas pageant was a delightful experience. 6. ^ The neighbors offered to chaparone when they heard that some teenagers might be having a party. 7. ^ The facts the politician presented seemed suspicious. 8. ^ Kenny was amazed to discover that his truckload of egges made the trip from Calgary to Vancouver intact. 9. ^ Kenley's friends felt that his contentment with welfare was a symptom of his distress. 10. ^ Rex couldn't notice any air pollution in New York, so he didn't worry about it. 11. ^ It was exhausting for Andrea to give so much time and effort to her classes. 12.^The older grandma got, the more she would bundle up to keep warm. 13. ^ It was great for Sally to find someone who shared her interest in baseball. 14. ^ The new game show host impressed the critics as highly talented. 15. ^ Society in the United States represents a blending of cultures. 16. ^ Cheeses have been developed with a wide variety of textures and flavors. 17. ^ It was felt that funds could be better administered under a single agency. 18. ^ All Einstein's ideas were spectacularly successful.  ^ ^  181 19. ^ Helen was surprised to hear that Pam and Max had gotten married. 20. ^ Black has been the color of dress for both the sophisticated and the rebellious. 21. ^ Cyclists are often surprised to find the highways in South America so wide and empty. 22. ^ The mechanics at Mike's Garage amused themselves with video movies. 23.^When Ellen married Peter, she had no idea that he was interested in her money. 24. ^ Her office mates felt that Karen was always looking for compliments on her new outfits. 25. ^ Tim was attractive, but his poverty was a liability he could not overcome. 26. ^ Jason was in demand as an interior decorator because he always catered to the tastes of his clients. 27. ^ Nigel thought Professor Rogers's theory was poorly thought out. 28. ^ The mountaineers set up six camps, one for each day of their climb. 29. ^ The lawyers were able to make a convincing case against the drug dealers. 30.^My Uncle Dave is always ready to help me with my problems. 31. ^ Corrine was hungry after a busy morning of classes. ^32. ^ Toby has always loved to read romance novels.  Metaphor Comprehension - 182  APPENDIX C  183  SENTENCE COMPREHENSION 6 Odie Geiger December, 1990 This experiment involves the reading and understanding of short groups of sentences. On each trial, a few sentences will be displayed, followed by a word to complete their meaning. Your task will be to read the sentences carefully, and when you are sure that you understand their meaning up to that point, push one of the buttons on the response box. Immediately, the sentences will disappear and the ending will be displayed. This portion of the experiment will be timed. As soon as you understand the ending so that the whole presentation makes sense, push the button again. A new set of sentences will then appear on the screen and the procedure will continue through 200 trials. It is important to the success of the experiment that you understand the meaning of the entire presentation. In order to ensure that you are reading the sentences carefully enough for your responses to be meaningful, a short quiz will follow the experiment. This does not mean that the sentences will be tricky or difficult. You will likely understand them as soon as you read them. Nor does it mean that you need to memorise the sentences, the details, or the names of the people who may be incidentally mentioned. It is just a way of keeping you from responding automatically as soon as the final words appear on the screen. Since the initial portion of the presentation is not timed, you can leave it on the screen as long as you like to make sure that its meaning is clear to you. After you press the button to display the ending, you should respond as quickly as possible with a second press of the button when the meaning of the whole phrase is understood. We will provide a few practice trials before the experiment begins to accustom you to the response procedure. You will have ten sets of 20 trials each to complete. If you become tired at any point, please feel free to take a break between sets. You will be given a point of class credit and an explanation of what the experiment is about when the session is over. Thank you for your interest and participation. Odie Geiger Graduate Student  DEADUNC3.APX^  184  EXPERIMENT 3 - DEAD METAPHORS 1)DD context: The recycling committee quickly got people to donate a steady supply of newspapers to its program. Unfortunately, finding a buyer for the papers was an unexpected 1)DD target: bottleneck. 1)DP target: difficulty. 1)DS target: expense. 1)DU target: impediment. 1)DL context: My grandmother has some old milk bottles in her basement. They were made to hold unhomogenized milk so that the cream would separate out and collect in the 1)DL target: bottleneck. 2)DD context: Tim's teachers were impressed with him from the start. He had the very appealing advantage of being both very personable and very 2)DD target: bright. 2)DP target: smart. 2)DS target: rich. 2)DU target: adroit. 2)DL context: After nearly two weeks of rain and overcast skies, Myrtle was shocked to look outside her window and see that the sky was clear and 2)DL target: bright. 3)DD context: Steve tried to establish a good relationship with his new mother-in-law, but despite all his efforts, she remained 3)DD target: cold. 3)DP target: indifferent. 3)DS target: shy. )DU target: aloof. 3)DL context: Alana loved the idea of going to study as an exchange student in Norway. All she knew about it was that it would be 3)DL target: cold. 4)DD context: Muriel and Howard were a contented couple most of the time, but when it came to politics, they each had opposing 4)DD target: positions. 4)DP target: opinions. 4)DS target: loyalties. 4)DU target: dogmas. 4)DL context: Joyce finally felt that she had achieved competence as a dancer when she realized how automatically she was able to assume the basic steps and 4)DL target: positions.  DEADUNC3.APX ^  185  5)DD context: All Roger knew anything about was cars and baseball. Sally felt that she had never before met anyone so 5)DD target: shallow. 5)DP target: superficial. 5)DS target: masculine. 5)DU target: insipid. 5)DL context: The beach at Lost Lake is perfect for young children. The sand extends for several meters and the water is warm and 5)DL target: shallow. 6)DD context: Warren realised that his problems were just getting worse as time went on. He wanted to make sense of his life, and spent years in a desperate search for a simple 6)DD target: formula. 6)DP target: solution. 6)DS target: faith. 6)DU target: rubric. 6)DL context: When Allison discovered a way to remove grass stains from clothing, several companies offered her large sums of money if she would sell them her 6)DL target: formula. 7)DD context: Whenever my problems seem too big to handle, I always know that I can turn to my Uncle Dave for 7)DD target: support. 7)DP target: comfort. 7)DS target: money. 7)DU target: succor. 7)DL context: One of the most important considerations for engineers is making sure that superstructures have adequate 7)DL target: support. 8)DD context: Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were 8)DD target: firm. 8)DP target: definite. 8)DS target: progressive. 8)DU target: unequivocal. 8)DL context: Cheeses have been developed with flavours varying from mild to tangy, and textures ranging from soft to 8)DL target: firm. 9)DD context: Tina preferred comedies or musicals to plays with social or political significance. She hated any form of entertainment that was too 9)DD target: deep. 9)DP target: profound. 9)DS target: expensive. 9)DU target: abstruse. 9)DL context: By the time the drillers actually found enough water to make pumping it worthwhile, the well was almost 50 meters 9)DL target: deep.  DEADUNC3.APX^  186  10)DD context: Don's chemistry class was invited to attend a lecture by a Nobel Prize winning French chemist. Don went, but he found the lecture difficult to 10)DD target: grasp. 10)DP target: understand. 10)DS target: translate. 10)DU target: comprehend. 10)DL context: After Bessie's illness, she was very weak. Her son helped 1 installing handrails throughout her house so that she would have something to 10)DL target: grasp. 11)DD context: Whenever Toby gets a new mystery novel by her favourite author, she puts everything else aside because she finds it so 11)DD target: absorbing. 11)DP target: interesting. 11)DS target: difficult. 11)DU target: engrossing. 11)DL context: The major oil companies are interested in developing a material for cleaning up oil spills that is harmless, yet highly oil 11)DL target: absorbing. 12)DD context: Many experienced cyclists are terrified of riding in South America, where roads are in poor repair and often have inadequate 12)DD target: shoulders. 12)DP target: edges. 12)DS target: restaurants. 12)DU target: verges. 12)DL context: As Grandma got older, she began to wear more warm clothes, even in the summertime. She never went out without putting a shawl over her 12)DL target: shoulders. 13)DD context: When Pam told Helen that she had eloped with Max against her parent's wishes and that they were now married, Helen was 13)DD target: shocked. 13)DP target: surprised. 13)DS target: happy. 13)DU target: appalled. 13)DL context: It is almost certain that if you touch an electrical appliance while you are in the bathtub you will get 13)DL target: shocked. 14)DD context: It was embarrassing for Jerry to forget his speech. Afterwards, his friends found him in a bar. They had never seen him so 14)DD target: depressed. 14)DP target: unhappy. 14)DS target: drunk. 14) DU -target: disheartened . 14)DL context: Even though we live in a rainy area, in the summer the water level at our municipal reservoir becomes dangerously 14)DL target: depressed. -  DEADUNC3.APX^  187  15)DD context: Rex liked the idea of living in New York but he worried about his health because the smell of industrial pollution was so 15)DD target: strong. 15)DP target: intense. 15)DS target: accepted. 15)DU target: pungent. 15)DL context: Casey chose his food carefully, exercised regularly, and took vitamin supplements in order to keep himself healthy and 15)DL target: strong. 16)DD context: Although she in many ways her life 16)DD target: parallel. 16)DP target: similar. 16)DS target: cursed. 16)DU target: allied. 16)DL context: It must have the original Canadian 16)DL target: parallel.  did not like to admit it, Bianca felt that and her mother's life were  been very difficult for the workers who built National Railway to keep the tracks exactly  17)DD context: The lawyers had assembled a highly detailed and carefully documented case against the suspected drug dealers. The judge felt the facts were 17)DD target: clear. 17)DP target: obvious. 17)DS target: unusual. 17)DU target: indisputable. 17)DL context: Bermuda is a great place to scuba dive. There is a huge variety of exotic tropical fish, and the water is warm and 17)DL target: clear. 18)DD context: Gordon was finally coming home after two years overseas. Sally felt calm, but when she finally spotted him at the airport, her feet began to 18)DD target: fly. 18)DP target: run. 18)DS target: itch. 18)DU target: jog. 18)DL context: The tiny robin was barely able to stand up when Theresa found him under a tree. It was a proud moment for both of them when he finally learned to 18)DL target: fly. 19)DD context: Nigel couldn't understand how Professor Rogers could get so much attention for his theory when his reasoning seemed so 19)DD target: weak. 19)DP target: vague. 19)DS target: silly. 19)DU target: imprec4aa, 19)DL context: The hostages were released after four years in captivity. They showed the effects of inadequate food and excercise, and were extremely 19)DL target: weak.  DEADUNC3.APX^  188  20)DD context: When Peter married Ellen, it was because she had enough  money to provide him with whatever he wanted. Ellen had no idea that he was so 20)DD target: calculating. 20)DP target: shrewd. 20)DS target: poor. 20)DU target: astute. 20)DL context: James always dreaded his monthly bookkeeping. He was never very good at math, and his balances often came out wrong even after hours of 20)DL target: calculating.  MODUNC3.APX^  189  EXPERIMENT 3 - METAPHORS 1)MM context: The committee decided to wait six months longer before  evaluating Laura's performance because they felt she was still target: green. target: inexperienced. target: recovering. target: unversed. context: Joanne thought that Ted would notice that she had painted their bedroom a different color, but to him it was still 1)ML target: green. 1)MM 1)MP 1)MS 1)MU 1)ML  2)MM context: David was highly regarded as an office manager. His success  came largely because he recognized that making an organization run smoothly required a lot of elaborate 2)MM target: choreography. 2)MP target: planning. 2)MS target: rules. 2)MU target: designing. 2)ML context: The dancers were thrilled to be able to work with someone who could offer them a program featuring such innovative 2)ML target: choreography.  3)MM context: Corrine had three classes that morning, with a quiz in one  and a presentation in another. By noon, she felt that she definitely needed to stop for 3)1414 target: fuel. 3)MP target: food. 3)MS target: exercise. 3)MU target: sustenance. 3)ML context: Art enjoyed the mental challenge of car rallies even more than the physical. One of his greatest delights was to keep a constant record of his car's consumption of 3)ML target: fuel. 4)MM context: When the producer invited Steve for an audition, he knew his  career might be determined in those few minutes. He wanted to do well, but he was aware that he was 4)MM target: wobbly. 4)MP target: nervous. 4)MS target: hiccuping. 4)MU target: flustered. 4)ML context: We knew how upset Mrs. Murray was when her cat got trapped in a tree, but we didn't expect her to rescue it herself, since she was so old and her ladder was so 4)ML target: wobbly.  MODUNC3.APX  ^  190  5)MM context: The United States represents a society in which the  cultures of a number of different nations have become 5)MM target: interwoven. 5)MP target: combined. 5)MS target: rebellious. 5)MU target: conjoined. 5)ML context: While travelling through India, Lois bought several meters of beautiful silk fabrics into which strands of metallic thread had been 5)ML target: interwoven. 6)MM context: It is important to realize that any creative endeavour  necessarily produces frustrations and mistakes. Not even Einstein's ideas were all 6)MM target: gems. 6)MP target: valuable. 6)MS target: original. 6)MU target: laudable. 6)ML context: Buying jewelry today is tricky. With so many synthetics available, the unwary buyer can easily be mislead into buying artificial materials, believing that they are 6)ML target: gems. 7)MM context: Carla told her psychiatrist that a full time job, plus a  husband, two children, and an elderly father at home was too much for her to 7)MM target: juggle. 7)MP target: manage. 7)MS target: conceal. 7)MU target: superintend. 7)ML context: When the stress of schoolwork made him tense, Charlie used to eat. He decided he needed a new pastime, and found exactly the diversion he needed when he learned to 7)ML target: juggle. 8)MM context: As the newest teacher on staff, Rebecca willingly accepted  responsibility for producing the school's Christmas pageant. By October, the project was becoming a 8)MM target: bear. 8)MP target: burden. 8)MS target: farce. 8)MU target: tribulation. 8)ML context: When my cousins from New York came to visit, the thing they wanted to do most was to camp out in the Rockies so they could photograph a live 8)ML target: bear.  MODUNC3.APX^  191  9)MM context: Walking into Margie's house is a retreat from the hectic,  noisy world. Margie is a true friend, someone whose nature is always pleasant, gentle, and 9)MM target: sunny. 9)MP target: cheerful. 9)MS target: simple. 9)MU target: genial. 9)ML context: Even though it was quite cold, we had a wonderful weekend of skiing. The snow stopped falling just after we arrived, and the skies became 9)ML target: sunny. 10)MM context: The politician's presentation was smooth, but the facts he presented were not consistent with what Terry knew to be true. She suspected the data had been 10)MM target: cooked. 10)MP target: changed. 10)MS target: stolen. 10)MU target: contrived. 10)ML context: Marilyn spent the morning in the orchard, picking apples for applesauce. She didn't stop working until evening, when the apples were finally peeled and 10)ML target: cooked.  11)MM context: After his team finished its second consecutive season at the very bottom of the league, the coach decided that his soccer team needed some new 11)MM target: blood. 11)MP target: players. 11)MS target: uniforms. 11)MU target: athletes. 11)ML context: Jerry wanted to be an ambulance attendant and help people. In order to get the job, he had to train himself to be insensitive to the sight of 11)ML target: blood. 12)MM context: Andrea tried to teach English, PE, and science to students  whose lives were only focused on personal concerns. Each night she went home feeling 12)MM target: empty. 12)MS target: sweaty. 12)MU target: ineffectual. 12)MP target: useless. 12)ML context: Paul hoped he would have a chance to talk to one of the speakers after the conference. He hurried back to the auditorium, but he found it 12)ML target: empty.  MODUNC3.APX^  192  13)MM context: Elsie, Bob, and Fluffy visited Julie at Christmas. When her brother phoned a week later to say that he and his dog would be arriving soon, Julie knew she couldn't stand another 13)MM target: attack. 13)MP target: guest. 13)MS target: animal. 13)MU target: caller. 13)ML context: When my father talks about his experiences in the war, I feel lucky that he's still alive. He often tells how just six soldiers were left to defend a whole village against an enemy 13)ML target: attack. 14)MM context: Carmen had always tried to think for herself. After she got too drunk at an office party, she wondered how she had been able to follow the 14)MM target: wave. 14)MP target: crowd. 14)MS target: tune. 14)MU target: throng. 14)ML context: The dangers of an earthquake are even greater for people living along a coastline who also have to worry about the possibility of a tidal 14)ML target: wave. 15)MM context: Edmund had written two novels before he turned 30. Later he wrote three more, but his editors rejected them all. They complained that his work had become 15)MM target: stagnant. 15)MP target: dull. 15)MS target: childish. 15)MU target: hackneyed. 15)ML context: It was shocking for Arnie to revisit his old home town after 20 years away. Ugly townhouses replaced open fields, and the pond where he often swam was now 15)ML target: stagnant. 16)MM context: It was time to call a strike. None of the changes that management proposed had been made, and the workers were starting to make 16)MM target: noises. 16)MP target: complaints. 16)MS target: mistakes. 16)MU target: innuendoes. 16)ML context: The mechanics at Mike's Garage amused one another by imitating the sounds customers would make as they tried to describe their car's odd 16)ML targets: noises.  MODUNC3.APX^  193  17)MM context: After the disastrous 1929 stock market crash, a lot of  formerly prosperous financiers saw their future prospects suddenly turn 17)MM target: black. 17)MP target: hopeless. 17)MS target: illegal. 17)MU target: glum. 17)ML context: Fashion is fickle. At different times, both sophisticated society and rebellious youth have focused their tastes around the color 17)ML target: black. 18)MM context: The studio audiences seemed to love the new game show host  but the critics and a number of people who watched the show felt that he was too 18)MM target: plastic. 18)MP target: artificial. 18)MS target: short. 18)MU target: affected. 18)ML context: One of the biggest problems faced by current environmentally conscious governments is what can be done to stop the production of throwaway 18)ML target: plastic. 19)MM context: Alice's playmates thought it was funny to pull the  beetle's legs off one by one and watch it struggle to walk, but Alice thought it was 19)MM target: sick. 19)MP target: horrible. 19)MS target: inspiring. 19)MU target: macabre. 19)ML context: It is often a difficult challenge for working couples to rearrange their day in order to get adequate home care when their child is 19)ML target: sick. 20)MM context: There has been a lot of inefficiency in our administration  of funds. Our new policy is to manage all of our funding under a single 20)MM target: umbrella. 20)14P target: agency. 20)MS target: bureaucrat. 20)MU target: federation. 20)ML context: It was a grey, rainy day and Harriet dozed on the bus until she reached her stop. After she had gotten off, she realized that she had forgotten her 20)ML target: umbrella.  ^  194  EXPERIMENT 3 SENTENCE COMPREHENSION 6 Odie Geiger name: ^ major: ^ class credit to (instructor): ^ Is English your most familiar language? ^ Please answer the following questions with a "T" if the meaning of the question is the same as that of one of the items you were given to read, or "F" if its meaning is different. Some of the items in the experiment were presented more than once with different endings; as long as the question s true of any ONE of those you remember, respond "T". ONce again, there are no trick questions here. For example, names are not switched between items, and incidents have not been combined or altered. 1.^Laura's committee recommended an immediate promotion for her. 2. ^ Tim was smart, but his rudeness tended to alienate his teachers. 3. ^ Alice's playmates tried to get the beetles to fight over bits of grain. 4. ^ All Alana knew about Norway was that it was a cold country. 5. ^ Rebecca found that producing the school's Christmas pageant was a delightful experience. 6. ^ It was David's policy to make plans that would keep his office functioning smoothly. Terry didn't trust the facts the politician prsented. 7. ^ Don provided his English speaking friends with a complete 8. ^ transcript of the French professor's talk. 9. ^ The recycling committee had trouble finding buyers for their collection of old newspapers. 10. ^ Rex couldn't notice any air pollution in New YOrk, so he didn't worry about it. 11. ^ It was exhausting for Andrea to give so much time and effort to her classes. The older grandma got, the more she would bundle up to keep 12. ^ warm. 13. ^ It was great for Sally to find someone who shared her interest in cars and baseball. 14. ^ The new game show host impressed the critics as highly talented. 15. ^ Society in the United States represents a blending of cultures. 16. ^ Cheeses have been developed with a wide variety of textures and flavors. 17. ^ It was felt that funds could be better administered under a single agency.  ^ ^  195 18. ^ All Einstein's ideas were spectacularly successful. 19. ^ Helen was surprised to hear that Pam and Max had gotten married. 20. ^ Black has been the color of dress for both the sophisticated and the rebellious. 21. ^ Cyclists are often surprised to find the highways in South America so wide and empty. 22. ^ The mechanics at Mike's Garage amused themselves with video movies. 23. ^ When Ellen married Peter, she had no idea that he was interested in her money. 24. ^ The coach was looking forward to leading his team into their third successful year. 25. ^ Julie had always looked forward to the visits by her family and friends. 26. ^ Tina's taste in drama tended towards Shakespeare and other dramatists of the Renaissance. 27. ^ Nigel thought Professor Roger's theory was poorly thought out. 28.^Jerry's friends rewarded his performance with a triumphant toast. 29. ^ The lawyers were able to make a convincing case against the drug dealers. 30.^My Uncle Dave is always ready to help me with my problems. 31. ^ Corrine was hungry after a busy morning of classes. 32. ^ Toby has always preferred to read science fiction novels.  Metaphor Comprehension - 196  APPENDIX D  ^  197  SENTENCE COMPLETION EXPERIMENT - OCTOBER. 1991 In this experiment, you will be given twenty different short contexts, usually one or two sentences long. The last word of the last sentence in each case will be missing. For each context, you are to list as many single, English words as you can think of that will complete the meaning of the sentence. TRY TO BE AS FLEXIBLE AND AS CREATIVE AS POSSIBLE. Nouns, verbs, adverbs, adjectives, metaphors, unusual or rare words are all acceptable, as long as (1) you have heard, read or used the word sometime before, (2) you know that you are using it correctly, and (3) the completed sentence makes sense to you. As an example, look over the following sample: a quiz in one ^that morning,^with Corrine had three classes and a presentation in another. By noon,^she felt that she definitely needed to stop for sustenance food a. exercise fuel g. b. coffee recharging h. c. a nap sandwiches d. prayer lunch j. e.  You will be allowed 2 minutes to come up with as many alternatives as you can. After this time, a beeper will sound to indicate that you are to go on to the next sentence. The screen beside you will inform you of which question number you should be on. Since this is an exercise in creativity, we would like to have you try to approach the sentences in as many ways as possible. Articles before words (a, the) are acceptable, but except for these, only single word completions are allowed. Try not to fill up spaces with synonyms of a single idea (coffee, tea, beer, juice, etc.) unless they express the idea in a different way (potables). If you think of an appropriate word for an earlier question after you have moved to a later one, write it down in the margin. After you finish, you can go back and add it to the list in the spaces labelled with an asterisk. Try as hard as you can to generate new ideas, but don't be disappointed it you can't fill all the allotted spaces in the time provided. This experiment should take about an hour, and you will receive one point of class credit. After the experiment is over, its purpose will be explained to you. Thank you for your interest and participation. e^ Gei g er, Research Assistant  198  1. The committee decided to wait six months longer before evaluating Laura's performance because they felt she was still ^ . a. ^  f. ^  b. ^  g. ^  c.^  h.  d. ^  1. ^  e  J. ^  . ^  2.Tim,s teachers were impressed with him from the start. He had the very appealing advantage of being both very personable and very ^ . a. ^ f. ^ b. ^  g. ^  c. ^  h. ^  d. ^  1. ^  e. ^  J. ^  it.^  *.  3. David was highly regarded as an office manager. His success came largely because he recognized that making an' organization run smoothly required a lot of elaborate a. ^  f. ^  b. ^  9. ^  c. ^  h. ^  d. ^  1. ^  e. ^  J. ^  *.^  *.  199  4. Steve tried to establish a good relationship with his new mother-in-law, but despite all his efforts, she remained a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  1. ^  e. ^  j. ^  5. It is important to realize that any creative endeavour necessarily produced frustrations and mistakes. Not even Einstein's ideas were all ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  1. ^  e. ^  j. ^  ^  *^  6.All Roger knew about anything about was cars and baseball. Sally felt that she had never before met anyone so^• a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  *  .^  *  200  7. Carla told her psychiatrist that a full time job, plus a husband, two children, and an elderly father at home was too much for her to ^ . a. b. ^  g. ^  c.^  h.  d. ^  I. ^  e.^  j.  8. Warren realised that his problems were just getting worse as time went on. He wanted to make sense of his life, and spent years in a desperate search for a simple ^. a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  I. ^  e. ^  i. ^  9. The politician's presentation was smooth, but the facts he presented were not consistent with what Terry knew to be true. She suspected the data had been ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  I. ^  e. ^  J. ^  201 10. Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were ^  a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. e. ^  j• ^  11. Andrea tried to teach English, PE, and science to students whose lives were only focused on personal concerns. Each night she vent home feeling ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. e. ^  j.  ^-  12. Tina preferred comedies or musicals to plays with social or political significance. She hated any form of ^• entertainment that was too ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  I. ^  e. ^  J. ^  202  13. Elsie, Bob, and Fluffy visited Julie at Christmas. When her brother phoned a week later to say that he and his dog would be arriving soon, Julie knew she couldn't stand another a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  i. ^  e. ^  j. ^  14. Don's chemistry class was invited to attend a lecture by a Nobel Prize winning French chemist. Don vent, but he found the lecture difficult to ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^ e. ^  j. ^  15. Edmund had written two novels before he turned 30. Later he wrote three more, but his editors rejected them all. They complained that his work had become a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  i. ^  e. ^  J. ^  *.^  *.  203  16. Whenever Toby gets a new mystery novel by her favourite author, she puts everything else aside because she finds it so a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^ e. ^  J. ^  17. The studio audiences seemed to love the new game show host, but the critics and a number of people who vatched the show felt that he was too ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^ e. ^  J. ^ *  18. When Pam told Helen that she had eloped vith Max against her parent's wishes and that they were nov married, Helen•was a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^ e. ^  J. ^  204  19. Alice's playmates thought it was funny to pull the beetle's legs off one by one and watch it struggle to walk, but Alice thought it was ^ a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^ e. ^  j. ^  20. Although she did not like to admit it, Bianca felt that in many ways her life and her mother's life were a. ^  f. ^  b. ^  g. ^  c. ^  h. ^  d. ^  I. ^  e.^  205 1.^The^committee^decided^to wait^six^months^longer^before evaluating Laura's performance because they felt she was still 1. active [1] 2. adjusting [1] 3. afraid [1] 4. agitated [1] 5. [an] amateur [1] 6. angry [1] 7. blossoming [1] 8. changing [1] 9. [a] convict [1] 10. dependent [1] 11. depressed [1] 12. developing [2] 13. doubtful [1] 14. early [1] 15. excellent [1] 16. exercising [1] 17. fat [1] 18. fearful [1] 19. frightened [1] 20. growing [2] 21. hurt [1] 22. hurting [1] 23. illiterate [1] 24. immature [2] 25. improving [3] 26. inadequate [1] 27. inappropriate [1] 28. incompetent [3] 29. incomplete [1] 30. inconsistent [1] 31. indecisive [1] 32. inept [1] 33. inexperienced [2] 34. injured [2] 35. insecure 1[] 36. irrational [1] 37. [a] klutz^[1] 38. learning [4] 39. [a] nerd^[1] 40. new [2] 41. nonconfident [1] 42. overweight [1] 43. practicing [4] 44. pregnant [1] 45. preparing [1] 46. psychotic [1] 47. queasy [1] 48. quiet [1] 49_ recouperating [1] 50. recovering [1]  51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75.  researching [1] serious [1] shocked [1] shy [1] sick^[3] single^[1] training [1] trying [1] unadapted [1] unable [1] anacceptable [1] uncommitted [1] underfed [1] uneasy [1] unfit^[1] unfoccused [1] unprepared [2] unqualified [1] unready [1] unsuitable [2] unsure [1] upset [1] winning [1] working [1] young [7]  76. in school [1] 77. not ready [1] 78. too serious [1]  206 2. Tim's teachers were impressed with him from the start. He had the very appealing advantage of being both very personable and very 1. active [1] 2. aggressive [1] 3. appealing [1] 4. artistic [1] 5. articulative^(articulate) [1] 6. assertive [1] 7. athletic [5] 8. attractive [3] 9. brave [1] 10. bright [3] 11. businesslike [1] 12. capable [1] 13. charismatic [1] 14. charming [1] 15. committed [1] 16. conscientious [1] 17. considerate [2] 18. cool [1] 19. cooperative [1] 20. courageous [1] 21. creative [4] 22. curious [1] 23. cute [2] 24. dedicated [1] 25. dependable [1] 26. direct [1] 27. eager [1] 28. efficient [2] 29. energetic [1] 30. entertaining [1] 31. enthusiastic [1] 32. erudite [1] 33. friendly [2] 34. funny [1] 35. gentle [1] 36. good-looking [1] 37. handsome [2] 38. hard-working [1] 39. helpful [4] 40. honest [4] 41. humourous [2] 42. independent [1] 43. informative [1] 44. ingenious [1] 45. intelligible [1] 46. intelligent [7] 47. interested [1] 4R. kind [1] 49. knowledgeable [1  50. likeable [1] 51. lovable [1] 52. masculine [1] 53. mature [1] 54. neat [1] 55. nice [2] 56. open [1] 57. open-minded [1] 58. optimistic [1] 59. organized [1] 60. original [1] 61. outgoing [2] 62. outspoken [2] 63. patient [1] 64. physical [1] 65. polite [4] 66. quiet [2] 67. respectable [1] 68. responsible [1] 69. rich [1] 70. sexy [1] 71. smart [2] 72. sociable [1] 73. soft-spoken [1] 74. studious [3] 75. sympathetic [1] 76. talkative [2] 77. tall^[1] 78. tardy [1] 79. uncompromising [1] 80. understanding [2] 81. vocal [1] 82. well-dressed [1] 83. witty [2] 84. young [1]  207 3. David was highly regarded as an office manager.^His success came largely because he recognized that making an organization run smoothly required a lot of elaborate 1. administrating [1] 2. advertising [1] 3. advisors [1] 4. architecture [1] 5. cases^[1] 6. changes [1] 7. clothes [1] 8. clothing [1] 9.commands [1] 10. committees [1] 11. communication [1] 12. companions [1] 13. computers [1] 14. concentration [1] 15. confrontations [1] 16. connections [2] 17. corrections [1] 18. creativity [1] 19. decisions [1] 20. decorations [1] 21. dedication [1] 22. delegation [1] 23. detailing [1] 24. details [1] 25. effort [1] 26. energy [1] 27. enthusiasm [1] 28. equipment [1] 29. expenditures [1] 30. experiences [1] 31. flattery [1] 32. forethought [1] 33. fortunetelling [1] 34. friends [1] 35. furniture [1] 36. goals [1] 37. graphs [1] 38. guesswork [1] 39. ideas^[4] 40. ingratiation [1] 41. judgements [1] 42. juggling [1] 43. kindness [1] 44. listening [1] 45. machinery [1] 46. machines [1] 47. management [1] 48. mediation [1] 49_ meetings [2] 50. networking [2]  51. orders [1] 52. organization [5] 53. orgies [1] 54. parties [1] 55. partners [1] 56. patience [1] 57. pens^[1] 58. people [1] 59. planning [11] 60. plans [2] 61. preaching [1] 62. precision [1] 63. presentations [1] 64. processes [1] 65. programming [1] 66. promotion [1] 67. relationships [1] 68. reports [1] 69. research [1] 70. schedules [1] 71. sense [1] 72. set-ups [1] 73. skill^[3] 74. spelling [1] 75. spirit [1] 76. stamina [1] 77. strategies [1] 78. studying [1] 79. talent [1] 80. teamwork [2] 81. thinking [2] 82. thought [2] 83. time^[1] 84. timing [1] 85. timetables [1] 86. training [1] 87. troubleshooting [1] 88. understanding [2] 89. work [4] 90. workers [2]  208 4. Stever tried to establish a good relationship with his new mother-in-law, but despite all his efforts, she remained 1. abusive [1] 2. accusing [1] 3. alcoholic [1] 4. alienated [1] 5. alive [1] 6. aloof [7] 7. ambivalent [1] 8. angry [2] 9. argumentative [1] 10. bitter [2] 11. brutal [1] 12. cold [6] 13. cold-eyed [1] 14. condescending [1] 15. cool [1] 16. despicable [1] 17. disagreeable [1] 18. disapproving [1] 19. discontent [1] 20. disinterested [2] 21. distant [5] 22. distrustful [1] 23. embarassing [1] 24. [an] enemy [1] 25. foreign [1] 26. frightening [1] 27. [a] gorilla [1] 28. hateful [1] 29. hostile [1] 30. ignorant [1] 31. impassive [2] 32. indifferent [2] 33. interfering [1] 34. intolerable [1] 35. irresponsible [1] 36. mean [1] 37. obnoxious [1] 38. [an] outsider [1] 39. pessimistic [2] 40. picky [1] 41. psychotic [1] 42. quiet [3] 43. rejecting [1] 44. removed [1] 45. reserved [1] 46. rude [1] 47. reclusive [1] 48. resentful [3] 49. sad [1] 50_ sadistic [a] 51. selfish [1]  52. self-righteous [1] 53. shy [1] 54. silent [1] 55. similarly [1] 56. sleazy [1] 57. [a] snob [1] 58. snobby [1] 59. speechless [1] 60. spiteful [1] 671. stubborn [2] 62. suspicious [1] 63. terrible [1] 64. unaccepting [1] 65. unamused [1] 66. unaffected [1] 67. unbearable [1] 68. uncooperative [1] 69. unconscious [1] 70. unfair [1] 71. unforgiving [1] 72. unfriendly [6] 73. unkind [1] 74. unpalatable [1] 75. unpersonable [1] 76. unpleasant [1] 77. unpleased [1] 78. unreachable [1] 79. unreasonable [1] 80. unsatisfied [2] 81. unsociable [1] 82. unsympathetic [1] 83. untouchable [1] 84. unwilling [1] 85. upset [2] 86. [a] witch [1] 87. intalkative [1]  ^  209 5. It^is^important^to^realize^that^any^creative^endeavor necessarily produces^frustrations^and^mistakes.^Not even Einstein's ideas were all 1. acceptable [2] 2. accepted [3] 3. adequate [1] 4. appreciated [1] 5. art [1] 6. believable [1] 7. brilliant [2] 8. Christian [1] 9. clear [2] 10. competent [1] 11. complete [1] 12. comprehensive [1] 13. concise [1] 14. consistent [1] 15. correct [12] 16. creative [1] 17. easy [2] 18. fantastic [1] 19. foolproof [1] 20. fulfilled [1] 21. fulfilling [1] 22. genius [1] 23. god-given [1) 24. good [2] 25. gorgeous [1] 26. great [1] • 27. his^[1] 28. honoured [1] 29. illuminating [1] 30. imaginative [1] 31. important [1] 32. ingenious [2] 33. intelligent [1] 34. interesting [1] 35. jokes^[1] 36. justifiable [1] 37. life-affirming [1] 38. logical [1] 39. normal [1] 40. noteworthy [1] 41. original [2] 42. perfect [6] 43. plausible [3] 44. popular [1] 45. productive [1] 46. proved [1] 47. proven [1] 48. rational [1] 49. rpadablem [1] 50. realized [1]  51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68.  reflective [1] responsible [1] right [3] sane^[2] sensible [1] specific [1] straightforward [2] sound [1] successful [1] true [1] truth [1] understandable [3] useful [3] valid [1] well-founded [1] well-liked [1] wise [1] workable [1]  69. great achievements [1]  210 6. All Roger knew anything about was cars and baseball. Sally felt she had never before met anyone so 1. amazing [1] 2. anal-retentive [1] 3. annoying [1] 4. athletic [1] 5. attached [1] 6. attractive [1] 7. backward [1] 8. boring [8] 9. chauvanistic [2] 10. childish [1] 11. compatible [1] 12. confined [1] 13. crazy [1] 14. dedicated [1] 15. determined [1] 16. distant [1] 17. dull^[3] 18. egocentrical [1] 19. engrossed [1] 20. entangled [1] 21. entertaining [2] 22. enveloped [1] 23. evil^[1] 24. exciting [2] 25. fixated [1] 26. focused [3] 27. gorgeous [1] 28. great [1] 29. grotesque [1] 30. homey [1] 31. ignorant [3] 32. immature [3] 33. incompatible [1] 34. incompetent [1] 35. incredible [1] 36. inexperienced [1] 37. intense [1] 38. interesting [3] 39. involved [1] 40. irresponsible [1] 41. knowledgeable [1] 42. lazy [1] 43. limited [1] 44. loyal^[1] 45. masculine [1] 46. matched [1] 47. mechanical [1] 48. narrow-minded [5] 49. obsessed [3] 50. obsellekte^[1] 51. old^[1]  52. outstanding [1] 53. primitive [1] 54. reclusive [1] 55. rednecked [1] 56. remarkable [1] 57. repetitive [1] 58. rugged [1] 59. selfish [1] 60. shallow [1] 61. similar [1] 62. simple [2] 63. single-minded [1] 64. smart [1] 65. special [1] 66. specialized [3] 67. splendid [1] 68. sporty [1] 69. square [1] 70. static [1] 71. stereotypical [2] 72. stupid [7] 73. submerged [1] 74. sure [1] 75. tall^[1] 76. typical [1] 77. unchallenging [1] 78. uneducated [3] 79. unimaginative [1] 80. inintelligent [1] 81. uninteresting [2] 82. unread [1] 83. useless [1] 84. wonderful [1] 85.  carried away [1]  211 7. Carla told her psychiatrist that a full-time job, plus a husbnd, two children, and an elderly father at home was too much for her to 1. abandon [1] 2. absorb [1] 3. accept [6] 4. analyze [1] 5. appreciate [2] 6. bear [1] 7. believe [1] 8. clean [1] 9. confront [1] 10. control [1] 11. cope [3] 12. do [3] 13. envision [1] 14. exercise [1] 15. expect [1] 16. experience [1] 17. explain [1] 18. feed [1] 19. finance [1] 20. grasp [1] 21. handle [14] 22. have [1] 23. imagine [1] 24. integrate [1] 25. juggle [1] 26. keep [2] 27. leave [1] 28. lose [1] 29. love [3] 30. maintain [1] 31. manage [2] 32. neglect [1] 33. ponder [1] 34. relax [2] 35. separate [1] 36. share [1] 37. smile [1] 38. stand [2] 39. take [3] 40. talk [1] 41. tolerate [1] 42. understand [1] 43. 44. 45. 46. 47. 48.  attend to [1] care about [1] cook for [1] cooperate with [1] concentrate on [1] listen to [1]  49. 50. 51. 52.  live with [1] look after [1] stress over [1] work with [1]  212 8. Warren realized that his problems were just getting worse as time went on.^Wh wanted to make some sense of his life, and spent years in a desperate search for a simple 1. answer [8] 2. belief^[1] 3. car [1] 4. cereal^[1] 5. clue [1] 6. companion [2] 7. cure [1] 8. deity [1] 9. drug [1] 10. end [2] 11. escape [1] 12. experiment [1] 13. explanation [1] 14. friend [1] 15. girl^[3] 16. goal [4] 17. hobby [1] 18. home [1] 19. house [1] 20. ideal^[1] 21. incentive [1] 22. island [1] 23. job^[7] 24. key [1] 25. life^[3] 26. lifestyle [4] 27. love [1] 28. meaning [2] 29. mentor [1] 30. path [2] 31. pet [1] 32. problem [1] 33. purpose [2] 34. question [2] 35. relationship [3] 36. religion [3] 37. rule [1] 38. satisfaction [1] 39. sleep [1] 40. solution [12] 41. suicide [1] 42. target [1] 43. task [1] 44. technique [1] 45. time [1] 46. toothpaste [1] 47. town [1] 48. treasure [1] 49. truth [3] 50. understanding [1]  51. vacation [1] 52. village [1] 53. wife^[1]  213 ^9.^The politician's presentation was smooth,^but the facts he presented were not consistent with what Terry knew to be true.^She suspected the data had been 1. added [1] 2. adjusted [1] 3. altered [4] 4. changed [2] 5. concocted [1] 6. conjured [1] 7. contrived [1) 8. controlled [1] 9. copied [1] 10. corrected [1] 11. created [2] 12. creative [1] 13. deleted [2] 14. discovered [1] 15. doctored [1] 16. engineered [1] 17. exaggerated [1] 18. fabricated [2] 19. fake [2] 20. faked [1] 21. false^[3] 22. falsified [5] 23. faulty^[1] 24. fictitious [1] 25. fixed [1] 26. forged [2] 27. forgotten [1] 28. fudged [1] 29. half-baked [1] 30. hallucinated [1] 31. imagined [2] 32. imaginary [1] 33. inadequate [1] 34. inconsistent [1] 35. incorrect [2] 36. invented [1] 37. inverted [1] 38. lies^[1] 39. limited [1] 40. made-up [2] 41. manipulated [2] 42. miscalculated [1] 43. misinterpreted [1] 44. misrepresented [1] 45. misunderstood [1] 46. mixed [1] 47. outdated [1] 48. overoptimistic [1] 49. plagerizcd [1] 50. questionable [1]  51. random [1] 52. recycled [1] 53. removed [1] 54. selected [1] 55. skewed [1] 56. stacked [1] 57. stolen [1] 58. switched [1] 59. tampered [4] 60. tainted [1] 61. unfounded [1] 62. unproven [1] 63. unreliable [1] 64. untrue [2] 65. worse [1] 66. wrong [4] 67. tampered with [1]  214 10. Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were 1. acceptable [1] 2. adequate [2] 3. appropriate [1] 4. approved [1] 5. bad[1] 6. [the] best [1] 7. changing [1] 8. continuing [1] 9. correct [4] 10. critical [1] 11. decided [2] 12. definite [1] 13. demanded [1] 14. different [1) 15. established [1] 16. excellent [1] 17. fair [1] 18. final (2) 19. finalized [1] 20. finished [1] 21. firm [2] 22. fixed [1] 23. foolproof [1] 24. good [2] 25. great [1] 26. ignored [1] 27. immutable [1] 28. implemented [1] 29. important [1] 30. indifferent [1] 31. inevitable [1] 32. inflexible (1) 33. intelligent [1] 34. irreversable [1] 35. justifiable [1] 36. justified [1] 37. law [1] 38. long-range [1] 39. made [2] 40. minimal [1] 41. necessary [3] 42. normal [1) 43. official [1] 44. overdue [1] 45. permanent [1] 46. positive [1] 47. practical [1] 48. proceeding [1] 50. proven [1]  51. reasonable [3] 52. responsible [1] 53. responsive [1] 54. right [3] 55. righteous [1] 56. rigid [1] 57. settled [1] 58. silly [1] 59. substantial [1] 60. superior [1] 61. supportive [1] 62. transferred [1] 63. unchangeable [2] 64. unstoppable [1] 65. untransferable [1] 66. useful [1] 67. useless [1] 68. wanted [1] 69. well-considered [1] 70. wise [1] 71. workable [1] 72. working [1] 73. going ahead [1] 74. in action [1] 75. in progress [1]  215 11. Andrea tried to teach English,^P.E. and science to students whose lives were only focused on personal concerns.^Each night she went home feeling 1. abused [1] 2. alone [2] 3. angry [7] 4. anxious [3] 5. astounded [1] 6. bad [1] 7. beseiged [1] 8. boring [1] 9. brilliant [1] 10. carefree [1] 11. concern [1] 12. confused [1] 13. content [1] 14. cranky [1] 15. crazy [1] 16. defeated [1] 17. dejected [1] 18. depressed [6] 19. detached [1] 20. disappointed [1] 21. discouraged [3] 22. disillusioned [1] 23. disturbed [1] 24. drained [3] 25. elated [1] 26. empty [1] 27. energetic [1] 28. enthusiastic [1] 29. excited [1] 30. exasperated [1] 31. exhausted [2] 32. fatigued [1] 33. frustrated [4] 34. gifted [1] 35. good [1] 36. guilty [1] 37. happy [4] 38. helpful [1] 39. helpless [1] 40. hopeful [1] 41. hopeless [2] 42. immature [1] 43. indecisive [1] 44. ineffective [1] 45. inspired [1] 46. intelligent [1] 47. lost [1] 48. mean [1] 49. meaningless [1] 50. miserable [1)  51. nice^[1] 52. nothing [1] 53. old^[1] 54. optimistic [1] 55. overworked [1] 56. pleased [1] 57. positive [1] 58. proud [1] 59. provocative [1] 60. redundant [1] 61. refreshed [1] 62. relaxed [1] 63. revived [1] 64. rejected [1] 65. sad^[1] 66. satisfied [6] 67. scared [1] 68. self-conscious [1] 69. self-righteous [1] 70. sick [1] 71. sleepy [1] 72. spiteful [1] 73. sorry [1] 74. stressful [1] 75. stupid [2] 76. superior [1] 77. tense [1] 78. tired [8] 79. uncertain [1] 80. unfulfilled [4] 81. unhappy [2] 82. unrewarded [1] 83. unwanted [1] 84. unworthy [1] 85. upset [1] 86. useful [2] 87. useless [3] 88. vulnerable [1] 89. wasted [1] 90. weird [1] 91. worried [1] 92. worthless [1] 93. worthwhile [1] 94. unaccomplishments [1]  216 12. Tina preferred comedies or musicals to plays with social or political significance.^She hated any form of entertainment that was too 1. abstract [1] 2. analytical [1] 3. artistic [1] 4. biased [2] 5. bland [1] 6. bored [1] 7. boring [8] 8. challenging [1] 9. complex [1] 10. conscientious [1] 11. conscious [1] 12. deep [2] 13. degrading [1] 14. delighted [1] 15. depressing [3] 16. didactic [2] 17. dramatic [2] 18. educational [1] 19. enlightening [1] 20. factual [1] 21. fixed [1] 22. focused [1] 23. general [1] 24. graphic [1] 25. heavy [3] 26. highbrow [1] 27. humourous [1] 28. imaginative [1] 29. immature [1] 30. implied [1] 31. inactive [1] 32. informative [1] 33. intellectual [4] 34. intelligent [2] 35. intense [1] 36. lifelike [1] 37. light-hearted [1] 38. mass-produced [1] 39. mature [1] 40. meaningful [3] 41. meaningless [1] 42. moral [1] 43. negative [1] 44. organized [1] 45. political [4] 46. pompouds [1] 47. predictable [1] 48. provocative [1] •  50. questionable [1]  51. real^[1] 52. realistic [6] 53. ridiculous [1] 54. sarcastic [1] 55. serious^[5] 56. silly^[1] 57. smart [1] 58. snobby [1] 59. stuffy [1] 60. stupid [1] 61. sub-standard [1] 62. substantial^[1] 63. stressing [1] 64. structured [1] 65. suggestive [1] 66. superficial [1] 67. symbolic [1] 68. tedious^[1] 69. thought-provoking [1] 70. tiring [1] 71. tricky [1] 72. true^[1] 73. unbelievable [1] 74. violent [2] 75. working-class [1] 76. worthwhile [1]  217 13. Elsie, Bob and Fluffy visited Julie at Christmas.^When her brother phoned a week later to say that he and his dog would be arriving soon, Julie knew she couldn't stand another 1. allergy [1] 2. animal [6] 3. arrival [1] 4. body [1] 5. bomb [2] 6. brother [1] 7. canine [1] 8. Christmas [2] 9. cleaning [1] 10. clean-up [1] 11. complaint [1] 12. compliment [1] 13. day [3] 14. delay [1] 15. disappointment [2] 16. disaster [1] 17. disruption [1] 18. disturbance [1] 19. dog [6] 20. event [1] 21. face^[1] 22. feast [1] 23. fight [2] 24. furball [1] 25. geek [1] 26. gift [2] 27. guest [3] 28. hamburger [1] 29. holiday [4] 30. hotdog [1] 31. hour [1] 32. houseguest [1] 33. inconvenience [1] 34. individual [1] 35. interference [1] 36. jerk [1] 37. joker [1] 38. licking [1] 39. male [1] 40. minute [3] 41. moment [1] 42. movie [1] 43. mutt [1] 44. news [1] 45. night [1] 46. noise^[1] 47. party [2] 48. person [6] • P 50. photograph [1]  .51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.  preparation [1] present [1] pressure [1] relative [8] responsibility [1] reunion [1] sandwich [1] shock [1] slob [1] statement [1] tackling [1] time [1] troublemaker [1] vacation [2] visit [7] visitor [4] voice [1] wait [1] week [2] word [1] year [1)  72. dog bone [1]  218 14. Don's chemistry class was invited to attend a lecture by a Nobel Prize winning French chemist. Don went, but he found the lecture difficult to 1. absorb [2] 2. appreciate [1] 3. attend [1] 4. awaken [1] 5. bear [1] 6. believe [3] 7. comply [1] 8. comprehend [9] 9. conceptualize [1] 10 . copy [1] 11. criticize [1] 12. enjoy [2] 13. explain [1] 14. find [1] 15. follow [7] 16. grasp [1] 17. hate [1] 18. hear [10] 19. hot [1] 20. identify [1] 21. incoherent [1] 22. intake [1] 23. interpret [3] 24. leave [1] 25. like [1] 26. listen [1] 27. question [1] 28. rationalize [1] 29. read [1] 30. realize [1] 31. relate [1] 32. scope [1] 33. see [3] 34. study [1] 35. stuffy [1] 36. support [1] 37. translate [3] 38. understand [14] 39. visualize [1] 40. 41. 42. 43. 44.  concentrate on [1] get into [1] listen to [1] relate to [3] write down [1]  219 15. Edmund had written two novels before he turned 30. Later he wrote^three^more,^but^his editors rejected^them^all.^They complained that his work had become 1. abusive [1] 2. aged [1] 3. amateurish [1] 4. arrogant [1] 5. bad [1] 6. boring [10] 7. childish [1] 8. commercialistic [1] 9. common [1] 10. confused [1] 11. crass^[1] 12. dreary [1] 13. drivel [1] 14. dry [1] 15. empty [1] 16. fictitious [1] 17. garbage [1] 18. gibberish [1] 19. idiotic [1] 20. inadequate [1] 21. inarticulate [1] 22. incoherent [2] 23. incomprehensible [1] 24. inconcise [1] 25. indecent [1] 26. inefficient [1] 27. infantile [1] 28. insignificant [1] 29. insufficient [1] 30. irrational [1] 32. irrelevant [1] 32. journalistic [1] 33. lengthy [1] 34. long-winded [1] 35. mainstream [1] 36. mature [2] 37. mediocre [1] 38. misleading [1] 39. moralizing [1] 40. obscene [1] 41. obvious [1] 42. old [1] 43. old-fashioned [1] 44. outdated [1] 45. outrageous [1] 46. passe [1] 47. pedantic [1] 48. pessimistic [1] 50. plain [1]  51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86.  political^[1] predictable [2] prejudice [1] realistic [1] repetitive [3] repititious [2] ridiculous [2] romantic [1] rusty [1] sadistic [1] serious [2] short [1] similar [1] simple [1] sloppy [4] stale [1] static [1] strange [1] stupid [1] stylized [1] tired [1] trite [1] unattractive [1] unbelievable [1] unenthusiastic [1] unethical [1] uneventful [1] unimportant [1] uninteresting [2] unorganized [1] unoriginal [1] unpopular [1] unreadable [1] unsellable [1] unsuitable [1] worst [1]  220 216.^Whenever Toby geta a new mystery novel by her favourite author,^she puts everything else aside because she finds it so  1. absorbing [4] 2. addicting [1] 3. alluring [1] 4. amusing [1] 5. appealing [2] 6. attracting [2] 7. attractive [1] 8. awesome [1] 9. boring [1] 10. challenging [2] 11. complex [2] 12. cool^[1] 13. creative [1] 14. descriptive [1] 15. different [1] 16. easy [1] 17. educational [1] 18. engaging [1] 19. engrossing [1] 20. enjoyable [1] 21. entertaining [5] 22. exciting [10] 23. exhilarating [1] 24. familiar [1] 25. fantastic [1] 26. fascinating [2] 27. fast-paced [1] 28. frightening [1] 29. fulfilling [1] 30. fun^[1] 31. funny [1] 32. good [1] 33. hilarious [2] 34. humourous [1] 35. hypnotic [1] 36. imaginative [2] 37. important [2] 38. inspirational [1] 39. intellectual [1] 40. intelligent [1] 41. intense [1] 42. interesting [11] 43. intriguing [5] 44. inviting [1] 45. irrelevant [1] 46. irresistible [2] 47. knowledgeable [1] 49. lovable [1]  50. meaningful [1] 51. mysterious [2] 52. pointless^[1] 53. provocative [1] 54. readable^[2] 55. real^[1] 56. realistic^[1] 57. relaxing [3] 58. rewarding [1] 59. romantic [1] 60. satisfying [1] 61. soothing [1] 62. spiritual^[1] 63. stimulating [1] 64. suspenseful [2] 65. tasteful^[1] 66. tempting [1] 67. thoughtful [1] 68. thrilling [1] 69. time-consuming [1] 70. tiring [2] 71. transient^[1] 72. worthwhile [1]  221 17. The studio audiences seemed to love the new game show host, but the critics and a number of people who watched the show felt that he was too 1. artificial [1] 2. attractive [1] 3. aggressive [4] 4. apologetic [1] 5. arrogant [1] 6. bald [1] 7. bizarre [1] 8. black [1] 9. blunt [1] 10. boisterous [1] 11. bored [1] 12. boring [3] 13. buoyant [1] 14. carefree [1] 15. chauvanistic [1] 16. childish [1] 17. common [1] 18. conservative [ 1 ] 19. crass [1] 20. critical [1] 21. crude [2] 22. depressing [2] 23. derogatory [1] 24. difficult [1] 25. doubtful [1] 26. dowdy [1] 27. dull^[2] 28. enthusiastic [2] 29. excited [1] 30. extreme [1] 31. fake [4) 32. false^[1] 33. fast-speaking [1] 34. fat^[5] 35. flashy [2] 36. fresh [1] 37. friendly [1] 38. funny [3] 39. giddy [1] 40. green [1] 41. gruff [1] 42. hairy [1] 43. handsome [2] 44. happy [3] 45. humourless [1] 46. humourous [1] 47. idealistic [1] 48. immature [1] 49. impatient [1] 50. incoherent [1]  51. indifferent [1] 52. irresponsible [1] 53. irritating [1] 54. laboured [1] 55. lecherous [2] 56. loud^[3] 57. loud-mouthed [1] 58. low-class [1] 59. manipulative [1] 60. mellow [1] 61. messy [1] 62. much [1] 63. nice^[1] 64. obnoxious [ 3 ] 65. old^[4] 66. outgoing [1] 67. outrageous [1] 68. outspoken [1] 69. overbearing [1] 70. personable [1] 71. personal [1] 72. pessimistic [2] 73. potbellied [1] 74. predictable [1] 75. quiet [2] 76. raucous [1] 77. remote [1] 78. repetitive [1] 79. rude [1] 80. ruthless [1] 81. sadistic [1] 82. sarcastic [1] 83. sensitive [1] 84. short [4] 85. showy [1] 86. shy [2] 87. sickly [1] 88. simple [1]d 89. skinny [1] 90. slick [1] 91. slim [1] 92. slow [2] 93. sly^[1] 94. smiley [1] 95. snooty [1] 96. stagnative [1] 97. stand-offish [1] 98. stereotypical [1] 99. stiff [1] 100. strange [1]  222 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118.  stupid [4] superficial [1] talkative [2] tall [3] ugly [4] unbelievable [1] unconventional [1] uncorteous [1] unkind [1] unoriginal [1] vague [1] violent [1] vocal [1] vociferous [1] weak [1] weird [1] wild [1] young [2]  223 18. When Pam told Helen that she had eloped with Max against her parent's wishes and that they were now married, Helen was 1. accepting [1] 2. afraid [2] 3. aghast [1] 4. aloof [1] 5. amused [2] 6. angry [13] 7. annoyed [1] 8. appalled [1] 9. astonished [3] 10. astounded [1] 11. baffled [1] 12. bathing [1] 13. bitter [1] 14. calm [1] 15. comforting [1] 16. confused [3] 17. concerned [2] 18. cooking [1] 19. crying [1] 20. curious [1] 21. depressed [1] 22. disappointed [1] 23. disapproving [1] 24. disgusted [2] 25. drunk [1] 26. eating [1] 27. ecstatic [3] 28. elated [2] 29. enchanted [1] 30. envious [3] 31. excited [4] 32. flabbergasted [3] 33. furious [2] 34. glad [2] 35. guilty [1] 36. happy [8] 37. helpful 1[] 38. honest [1] 39. horrified [1] 40. humoured [1] 41. hungry [1] 42. hurt [1] 43. interested [1] 44. itchy [1] 45. jealous [5] 46. joyous [1] 47. judgemental [1] 48. impressed [1] 49. intrigued [1] 50. kind [1] 51. mad [1]  52. mean [1] 53. melancholy [1] 54. moralizing [1] 55. mortified [1] 56. mystified [1] 57. negative [1] 58. nervous [1] 59. optimistic [1] 60. outraged [1] 61. overwhelmed [1] 62. preoccupied [1] 63. proud [1] 64. relieved [3] 65. sad [1] 66. scared [2] 67. shocked [11] 68. sick [1] 69. smiling [1] 70. sorry [1] 71. speechless [3] 72. stunned [1] 73. supportive [1] 74. surprised [7] 75. tearful [1] 76. terrified [1] 77. triumphant [1] 78. understanding [2] 79. unforgiving [1] 80. unsure [1] 81. unsupportive [1] 82. upset [3] 83. violent [1] 84. warm [1] 85. worried [4] 86. [a] wreck [1]  224 19. Alice's playmates thought it was funny to pull the beetle's legs off one by one and watch it struggle to walk, but Alice thought it was 1. abnormal [1] 2. admirable [1] 3. appalling [1] 4. awful [3] 5. bleeding [1] 6. boring [6] 7. brilliant [1] 8. childish [2] 9. cruel [13] 10. dead [1) 11. degrading [1] 12. depressing [1] 13. disgusting [7] 14. disturbing [1] 15. [a] dream [1] 16. dying [1] 17. escaping [1] 18. fascinating [2] 19. funnier [1) 20. funny [1] 21. gross [3] 22. harmful [1] 23. hilarious [2] 24. horrible [2] 25. horrifying [1] 26. hysterical [1] 27. illuminating [1) 28. immature [3] 29. immoral [1] 30. inappropriate [1] 31. inhuman[e] [6] 32. kind [1] 33. mean [4] 34. meaningless [1] 35. monstrous [1] 36. metamorphosing [1] 37. neat [1] 38. nice [1] 39. non-human [1] 40. normal [1] 41. not [1] 42. painful [1] 43. right [1] 44. [a] ritual [1] 45. sad [3] 46. saddening [1] 47. sadistic [3] 48. savagely [1] 49. sick [2] 50. spell-binding [1)  51. [a] sport [1] 52. strange [2] 63. stupid [5] 54. suffereing [1] 55. terrible [1] 56. tired [1] 57. torture [1] 58. tyrannical [1] 59. unattractive [1] 60. unbelievable [1] 61. uncaring [1] 62. uncivilized [1] 63. uneventful [1] 64. unfair [1] 65. ungodly [1] 66. unhumanistic [1] 67. unoriginal [1] 68. unusual [1) 69. vengeful [1] 70. violent [2] 71. warped [1] 72. [a] waste [1] 73. weird [1] 74. wrong [1] 75. a waste of time [1]  225 20. Although she did not like to admit it, Bianca felt that in many ways her life and her mother's life were ^. 1. abnormal [2] 2. active [1] 3. associated [1] 4. beautiful [1] 5. [the] best [1] 6. bizarre [1] 7. boring [6] 8. bounded [1] 9. calm [1] 10. chaotic [1] 11. colorful^[1] 12. connected [2] 13. contrastive [1] 14. controlled [1] 15. convergent [1] 16. creative [1] 17. dangerous [3] 18. dependent [1] 19. different [9] 20. disparate [1] 21. distant [1] 22. diverging [1] 23. doomed [1] 24. endangered [1] 25. empty [1] 26. ending [1] 27. entertaining [1] 28. equal [1] 29. exciting [2] 30. extreme [1] 31. failures^[1] 32. frightening [1] 33. frivolous [1] 34. fulfilled [1] 35. fulfilling [1] 36. fun [1] 37. funny [1] 38. glamourous [1] 39. happy [2] 40. hard [1] 41. hectic [1] 42. hopeless [2] 43. identical^[2] 44. incompatible [1] 45. independent [1] 46. inseparate [1] 47. intellectual [1] 48. interconnected [1] 49. intertwined [1] cO_ intransigent [1] 51. intriguing [1)  52. limited [1] 53. linked [1] 54. lively [1] 55. lonely [2] 56. meaningless [1] 57. opposite [1] 58. over [1] 59. overlapping [1] 60. pathetic [1] 61. parallel^[1] 62. perfect [1] 63. poor [1] 64. promising [1] 65. repeating [1] 66. restricted [1] 67. routine [1] 68. routined [1] 69. ruined [1] 70. sad^[3] 71. [the] same [3] 72. satisfactory [1] 73. sedentary [1] 74. selfish [1] 75. separate [2] 76. separated [1] 77. short [1] 78. similar^[12] 79. spoiled [1] 80. spontaneous [1] 81. steady [1] 82. stressful [1] 83. superficial^[1] 84. supressed [1] 85. tedious [1] 86. terrible [1] 87. together [1] 88. typical [1] 89. unalike [1] 90. unconventional [1] 91. unfulfilled [1] 92. uninteresting [2] 93. unpredictable [1] 94. unusual [2] 95. worthless [1]  Metaphor Comprehension - 226  APPENDIX E  227  LEXICAL DECISION - APPROPRIATE VS. INAPPROPRIATE - SENLEX June, 1992 This experiment consists of a series of short sentence contexts in which the final word of the final sentence is missing. Your task will be to choose which of two words completes the intended meaning of the sentence. After making your choice, we want you to indicate, first for the chosen word and then for the rejected word, just how appropriate you think these two choices are to the sentence context. Using a scale like the one below, give a "1" to the words you judge to be highly appropriate, and a "7" to the ones you believe are absolutely inappropriate; the ones that make no sense at all. 'Appropriate' means a word that might actually be used to complete the meaning of the sentence context, not just one that is somehow related to the topic. A word can be highly appropriate even it is surprising or unfamiliar in terms of the expectations generated by the sentences. Do not worry if you find yourself using some categories more than others, this is not at all unlikely. Just try to give your honest first response to each sentence in terms of its ending. 1^2^3^4^5^6^7 highly^fairly^fairly^highly appropriate^appropriate^inappropriate^inappropriate There will likely be some words in the experiment that are unfamiliar to you, or you feel you understand only vaguely. This may be enough to guide your decision as to which word to choose, but leave you unable to assess its appropriateness. If you feel you are really too unfamiliar with the word to judge whether it is appropriate or not, please put an "X" in that space. It is important not to confuse unfamiliarity with inappropriateness - that is. do not call a word inappropriate only because you are unfamiliar with it. The sentence contexts will be displayed all at once on a computer monitor. Most are one to two sentences long. You are to read the sentences, and once you have understood their meaning, press the bottom two buttons on the box as indicated. This will remove the context and present you with two words, side by side, on the video screen. Read them both, and decide which of the two is the most appropriate ending for the context. Then think about how appropriate you feel it is, and rank it on the paper according to the 7point scale. Do the same for the less appropriate word on the second line. The top line for each air should alwa s be a lower num Pi 1 • 1 1 " • 11  228  line. Once you have made your decision and rated both words, press the button on the top of the button box corresponding to the word that you have rated first (the more appropriate word). A new context will appear and the rating procedure should continue through 200 trials. We want you to respond as accurately as possible, but the decisions should be fairly spontaneous. No decision should take you more than a few seconds. You will receive one class credit for about an hour of participation. Thank you for your interest. Odie Geiger, Research Assistant Dr. Lawrence Ward, Supervising Professor  229  LEXICAL DECISION - APPROPRIATE VS. INAPPROPRIATE - SENLEX June, 1992 Please rank the more appropriate of the two words in the "a" space and the less appropriate in the 'b" space. Give a "1" to the words you judge to be highly appropriate, and a "7" to the ones you believe are absolutely inappropriate. If you feel you are too unfamiliar with the word to judge whether it is appropriate or not, please put an "X" in that space. It is important not to confuse unfamiliarity with inappropriateness - that is. do not call a word inappropriate only because you are unfamiliar with it. 1  ^  2^3^4^5^6^7  highly^fairly^fairly highly appropriate^appropriate^inappropriate^inappropriate ************************** PRACTICE ***************************** la. 5a. 5b. lb. 2a. 2b. 3a. 3b. 4a. 4b. ***************************** SET 1 ***************************** 13a. 17a. la. 9a. 5a. 13b. 9b. 17b. lb. 5b. 2a. 2b.  6a. 6b.  10a. 10b.  14a. 14b.  18a. 18b.  3a. 3b.  7a. 7b.  lla. 11b.  15a. 15b.  19a. 19b.  4a. 4b.  8a. 8b.  12a. 12b.  16a. 16b.  20a. 20b.  230  SET 2 ***************************** la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. * * * * * * * * * * * * * * * * * * * * *** ** ***  2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a.^7a.^lla.^15a.^19a. 3b.^7b.^11b.^15b.^19b. 4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  *** * * * *^* **** * ** *********** SET 3 ******'************************  la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. 2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a. 3b.  7a.^lla.^15a.^19a. 7b.^11b.^15b.^19b.  4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  ***************************** S ET 4 ***************************** la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. 2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a. 3b.  7a.^lla.^15a.^19a. 7b.^11b.^15b.^19b.  4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  231  * * * * * * * * * * * * * * * * * * ********** * SET 5 *****************************  la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. 2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a. 3b.  7a.^lla.^15a.^19a. 7b.^11b.^15b.^19b.  4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  ******* ** ** ****** * ********* SET  6 *****************************  la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. 2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a.^7a.^lla.^15a.^19a. 3b.^7b.^11b.^15b.^19b. 4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  ***************************** SET 7 ***************************** la.^5a.^9a.^13a.^17a.  lb.^5b.^9b.^13b.^17b. 2a. 2b.  ^ 6a. 10a.^14a.^18a. ^ 10b.^14b.^18b. 6b.  3a. 3b.  7a.^I la.^15a.^19a. 7b.^11b.^15b.^19b.  4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  232  * * * * * * * * * * * * * * * * * * * ********* SET  8 ******************************  la.^5a.^9a.^13a.^17a. lb.^5b.^9b.^13b.^17b. 2a. 2b.  6a.^10a.^14a.^18a. 6b.^10b.^14b.^18b.  3a. 3b.  7a.^lla.^15a.^19a. 7b.^11b.^15b.^19b.  4a. 4b.  8a.^12a.^16a.^20a. 8b.^12b.^16b.^20b.  **************************** SET 9 ******************************* la. lb.  5a. 5b.  9a. 9b.  13a. 13b.  17a. 17b.  2a. 2b.  6a. 6b.  10a. 10b.  14a. 14b.  18a. 18b.  3a. 3b.  7a. 7b.  11a. 11b.  15a. 15b.  19a. 19b.  4a. 4b.  8a. 8b.  12a. 12b.  16a. 16b.  20a. 20b.  ***************************** SET  *****************************  la. lb.  5a. 5b.  9a. 9b.  13a. 13b.  17a. 17b.  2a. 2b.  6a. 6b.  10a. 10b.  14a. 14b.  18a. 18b.  3a. 3b.  7a. 7b.  lla. 1 lb.  15a.  15b.  19a.  19b.  4a. 4b.  8a. 8b.  12a. 12b.  16a. 16b.  20a. 20b.  233  LEXICAL DECISION - APPROPRIATE VS. INAPPROPRIATE - SENLEX June, 1992 This experiment consists of a series of short sentence contexts in which the final word of the final sentence is missing. Your task will be to choose which of two words completes the intended meaning of the sentence, and then to indicate your choice by pressing a button corresponding to the chosen word as quickly as possible. The sentence contexts will be displayed all at once on a computer monitor. Most are one or two sentences long. You are to read the sentences, and once you have understood their meaning, press the bottom buttons on the box as indicated. This will remove the context and present you with two words, side by side, on the video screen. Read them both, and then press either the right or left button on the top of the box corresponding to the position of the word. Once you have made your response, a new sentence context will automatically appear on the screen. As the presentation continues, you will notice that some of the contexts are repeated. However, the endings for these sentences will be different. It is important to the success of this experiment that you read through every context to the point of understanding each time it appears, even if you recall it from a previous display. There will be a short pencil and paper quiz following the experiment that you will find easy to answer as long as each context is understood in terms of its endings. If you are ever unfamiliar with the meaning of one or both of the words presented, or if you are unable to decide between the two, or if neither of the words seems to make sense to you in terms of the context, just make a guess and continue on with the experiment. You should try to work at a steady pace, and to be as accurate as possible in making your choice. The entire experiment consists of 200 trials of sentence contexts followed by the quiz, and should take about one hour to complete. You will receive one hour of class credit for your participation. Thank you for your interest. Odie Geiger, Research Assistant Dr. Lawrence Ward, ^Supervising  DEADUNC5.APX  ^  234  EXPERIMENT 5 - DEAD METAPHORS Note: In this appendix, correct targets are listed first, followed by matched incorrect targets. In the actual experiment, the correct stimulu was randomly selected to appear on either the right or left side of the screen. 1)DD context: The recycling committee quickly got people to donate a steady supply of newspapers to its program. Unfortunately, finding a buyer for the papers was an unexpected 1)DD target: bottleneck^sketchbook 1)DP target: difficulty^atmosphere 1)DS target: expense^welcome 1)DU target: impediment^hypodermic 1)DL context: My grandmother has some old milk bottles in her basement. They were made to hold unhomogenized milk so that the cream would separate out and collect in the 1)DL target: bottleneck^paraphrase 2)DD context: Tim's teachers were impressed with him from the start. He had the very appealing advantage of being both very personable and very 2)DD target: bright^battle 2)DP target: smart^valid 2)DS target: rich^gain 2)DU target: astute^malted 2)DL context: After nearly two weeks of rain and overcast skies, Myrtle was shocked to look outside her window and see that the sky was clear and 2)DL target: bright^weight 3)DD context: Steve tried to establish a good relationship with his new mother-in-law, but despite all his efforts, she remained 3)DD target: cold^move 3)DP target: indifferent^seventeenth 3)DS target: shy^par 3)DU target: aloof^plaid 3)DL context: Alana loved the idea of going to study as an exchange student in Norway. All she knew about it was that it would be 3)DL target: cold^sort 4)DD context: Muriel and Howard were a contented couple most of the time, but when it came to politics, they each had opposing 4)DD target: positions^elections 4)DP target: opinions^supplies 4)DS target: loyalties^footnotes 4)DU target: stances^oatmeal 4)DL context: Joyce finally felt that she had achieved competence as a dancer when she realized how automatically she was able to assume 4)DL target: positions ^employees  DEADUNC5.APX  235  5)DD context: All Roger knew anything about was cars and baseball. Sally felt that she had never before met anyone so 5)DD target: shallow^pasture 5)DP target: superficial^microscopic 5)DS target: masculine^scratched 5)DU target: insipid^twirled 5)DL context: The beach at Lost Lake is perfect for young children. The sand extends for several meters and the water is warm and 5)DL target: shallow ^amended 6)DD context: Warren realised that his problems were just getting worse as time went on. He wanted to make sense of his life, and spent years in a desperate search for a simple 6)DD target: formula^tension 6)DP target: solution^railroad 6)DS target: faith^staff 6)DU target: tenet^snout 6)DL context: When Allison discovered a way to remove grass stains from clothing, several companies offered her large sums of money if she would sell them her 6)DL target: formula^housing 7)DD context: Whenever my problems seem too big to handle, I always know that I can turn to my Uncle Dave for 7)DD target: support^looking 7)DP target: comfort^regions 7)DS target: money^death 7)DU target: sustenance^casualties 7)DL context: One of the most important considerations for engineers is making sure that superstructures have adequate 7)DL target: support^picture 8)DD context: Angry parents wanted the government to stop cutbacks to schools, but the Department of Education maintained that their plans were 8)DD target: firm^cars 8)DP target: definite^swimming 8)DS target: progressive ^respirtory 8)DU target: unequivocal ^chauffeured 8)DL context: Cheeses have been developed with flavours varying from mild to tangy, and textures ranging from soft to 8)DL target: firm^pool 9)DD context: Tina preferred comedies or musicals to plays with social or political significance. She hated any form of entertainment that was too 9)DD target: deep^unit 9)DP target: profound^released 9)DS target: expensive^separated g •  9)DL context: By the time the drillers actually found enough water to make pumping it worthwhile, the well was almost 50 meters 9)DL target: deep^jazz  DEADUNC5.APX ^  236  10)DD context: Don's chemistry class was invited to attend a lecture by a Nobel Prize winning French chemist. Don went, but he found the lecture difficult to 10)DD target: grasp^tumor 10)DP target: understand^population 10)DS target: translate^dimension 10)DU target: comprehend^accelerate 10)DL context: After Bessie's illness, she was very weak. Her son helped installing handrails throughout her house so that she would have something to 10)DL target: grasp^quote 11)DD context: Whenever Toby gets a new mystery novel by her favourite author, she puts everything else aside because she finds it so 11)DD target: absorbing^shuffling 11)DP target: interesting^responsible 11)DS target: difficult^committee 11)DU target: engrossing ^unsinkable 11)DL context: The major oil companies are interested in developing a material for cleaning up oil spills that is harmless, yet highly of 11)DL target: absorbing^threading 12)DD context: Many experienced cyclists are terrified of riding in South America, where roads are in poor repair and often have inadequate 12)DD target: shoulders^decisions 12)DP target: edges^knees 12)DS target: restaurants^metaphysics 12)DU target: verges^rivals 12)DL context: As Grandma got older, she began to wear more warm clothes, even in the summertime. She never went out without putting a shawl over her 12)DL target: shoulders^elections 13)DD context: When Pam told Helen that she had eloped with Max against her parent's wishes and that they were now married, Helen was 13)DD target: shocked ^blowing 13)DP target: surprised^organized 13)DS target: happy^glass 13)DU target: appalled ^prolific 13)DL context: It is almost certain that if you touch an electrical appliance while you are in the bathtub you will get 13)DL target: shocked^broader 14)DD context: It was embarrassing for Jerry to forget his speech. Afterwards, his friends found him in a bar. They had never seen him so 14)DD target: depressed rendering 14)DP target: unhappy stained 14)DS target: drunk magic 14)DU target: disheartened translusccnt 14) DL context: Even though we live in a rainy area, in the summer the water level at our municipal reservoir becomes dangerously 14)DL target: depressed^respected  DEADUNC5.APX  ^  237  15)DD context: Rex liked the idea of living in New York but he worried about his health because the smell of industrial pollution was so 15)DD target: strong^beyond 15)DP target: intense^concert 15)DS target: accepted^opposite 15)DU target: pungent^tubular 15)DL context: Casey chose his food carefully, exercised regularly, and took vitamin supplements in order to keep himself healthy and 15)DL target: strong^inside 16)DD context: Although she did not like to admit it, Bianca felt that in many ways her life and her mother's life were 16)DD target: parallel ^regional 16)DP target: similar^earlier 16)DS target: cursed^porous 16)DU target: akin^surf 16)DL context: It must have been very difficult for the workers who built the original Canadian National Railway to keep the tracks exactly 16)DL target: parallel^enormous 17)DD context: The lawyers had assembled a highly detailed and carefully documented case against the suspected drug dealers. The judge felt the facts were 17)DD target: clear^north 17)DP target: obvious^musical 17)DS target: unusual^highest 17)DU target: indisputable^aeronautical 17)DL context: Bermuda is a great place to scuba dive. There is a huge variety of exotic tropical fish, and the water is warm and 17)DL target: clear^alone 18)DD context: Gordon was finally coming home after two years overseas. Sally felt calm, but when she finally spotted him at the airport, her feet began to 18)DD target: fly ^pat 18)DP target: run ^cut 18)DS target: itch ^slam 18)DU target: jog^rib 18)DL context: The tiny robin was barely able to stand up when Theresa found him under a tree. It was a proud moment for both of them when he finally learned to 18)DL target: fly^bid 19)DD context: Nigel couldn't understand how Professor Rogers could get so much attention for his theory when his reasoning seemed so 19)DD target: weak ^hurt 19)DP target: vague ^brave 19)DS target: silly^lined 19)DU target: imprecise ^sleepless 19) DL cvuteXL. TLe hustayes were rel eased aft er four years in captivity. They showed the effects of inadequate food and excercise, and were extremely 19)DL target: weak^bare  DEADUNC5.APX  ^  238  context: When Peter married Ellen, it was because she had enough money to provide him with whatever he wanted. Ellen had no idea that he was so 20)DD target: calculating^fragmentary 20)DP target: shrewd^barred 20)DS target: poor^grey 20)DU target: wily^dank 20)DL context: James always dreaded his monthly bookkeeping. He was never very good at math, and his balances often came out wrong even after hours of 20)DL target: calculating^adolescents 20)DD  MODUNC5.APX  ^  239  EXPERIMENT 5 - METAPHORS Note: In this appendix, correct targets are listed first, followed by matched incorrect targets. In the actual experiment, the correct stimuli was randomly selected to appear on either the right or left side of the screen. 1)MM context: The committee decided to wait six months longer before evaluating Laura's performance because they felt she was still 1)MM target: green^seven 1)MP target: inexperienced ^environmental 1)MS target: recovering^synonomous 1)MU target: unversed^embossed 1)ML context: Joanne thought that Ted would notice that she had painted their bedroom a different color, but to him it was still 1)ML target: green^teeth 2)MM context: David was highly regarded as an office manager. His success came largely because he recognized that making an organization run smoothly required a lot of elaborate 2)MM target: choreography^astrophysics 2)MP target: planning^religion 2)MS target: rules^youth 2)MU target: designing ^figurines 2)ML context: The dancers were thrilled to be able to work with someone who could offer them a program featuring such innovative 2)ML target: choreography^bewilderment 3)MM context: Corrine had three classes that morning, with a quiz in one and a presentation in another. By noon, she felt that she definitely needed to stop for 3)MM target: fuel^sums 3)MP target: food^hair 3)MS target: exercise^machines 3)MU target: sustenance^highlights 3)ML context: Art enjoyed the mental challenge of car rallies even more than the physical. One of his greatest delights was to keep a constal record of his car's consumption of 3)ML target: fuel^cats 4)MM context: When the producer invited Steve for an audition, he knew hi: career might be determined in those few minutes. He wanted to do well, but he was aware that he was 4)MM target: wobbly^clawed 4)MP target: nervous^biggest 4)MS target: hiccuping^garlanded 4)MU target: flustered^vexatious 4)ML context: We knew how upset Mrs. Murray was when her cat got trapped in a tree, but we didn't expect her to rescue it herself, since she was so old and her ladder was so 4)ML target: wobbly^pitied  MODUNC5.APX  ^  240  5)MM context: The United States represents a society in which the cultures of a number of different nations have become 5)MM target: interwoven ^morticians 5)MP target: combined^hydrogen 5)MS target: rebellious^bequeathed 5)MU target: conjoined^subtitled 5)ML context: While travelling through India, Lois bought several meters of beautiful silk fabrics into which strands of metallic thread had been 5)ML target: interwoven ^reconvened 6)MM context: It is important to realize that any creative endeavour necessarily produces frustrations and mistakes. Not even Einstein's ideas were all 6)MM target: gems^pegs 6)MP target: valuable^changing 6)MS target: original^returned 6)MU target: laudable^buttoned 6)ML context: Buying jewelry today is tricky. With so many synthetics available, the unwary buyer can easily be mislead into buying artificial materials, believing that they are 6)ML target: gems^hubs 7)MM context: Carla told her psychiatrist that a full time job, plus a husband, two children, and an elderly father at home was too much for her to 7)MM target: juggle^dilute 7)MP target: manage^bundle 7)MS target: conceal^breathe 7)MU target: superintend^firecracker 7)ML context: When the stress of schoolwork made him tense, Charlie used to eat. He decided he needed a new pastime, and found exactly the diversion he needed when he learned to 7)ML target: juggle^staple 8)MM context: As the newest teacher on staff, Rebecca willingly accepted  responsibility for producing the school's Christmas pageant. By October, the project was becoming a 8)MM target: bear^veto 8)MP target: burden^lawyer 8)MS target: farce^leash 8)MU target: tribulation^requisition 8)ML context: When my cousins from New York came to visit, the thing they wanted to do most was to camp out in the Rockies so they could photograph a live 8)ML target: bear^rust  MODUNC5.APX^  9)MM context: Walking into Margie's house is a retreat from the noisy world. Margie is a true friend, someone whose nature always pleasant, gentle, and 9)MM target: sunny^blade 9)MP target: cheerful^deceased 9)MS target: simple^higher 9)MU target: genial^dotted 9)ML context: Even though it was quite cold, we had a wonderful of skiing. The snow stopped falling just after we arrived, skies became 9)ML target: sunny^aloud  241  hectic, is  weekend and the  10)MM context: The politician's presentation was smooth, but the facts he presented were not consistent with what Terry knew to be true. She suspected the data had been 10)MM target: cooked^caring 10)MP target: changed ^talking 10)MS target: stolen^cosmic 10)MU target: contrived^nocturnal 10)ML context: Marilyn spent the morning in the orchard, picking apples for applesauce. She didn't stop working until evening, when the apples were finally peeled and 10)ML target: cooked ^reared 11)MM context: After his team finished its second consecutive season at the very bottom of the league, the coach decided that his soccer team needed some new 11)MM target: blood^trees 11)MP target: players ^samples 11)MS target: uniforms^feathers 11)MU target: athletes^excerpts 11)ML context: Jerry wanted to be an ambulance attendant and help people. In order to get the job, he had to train himself to be insensitive to the sight of 11)ML target: blood^truth 12)MM context: Andrea tried to teach English, PE, and science to students whose lives were only focused on personal concerns. Each night she went home feeling 12)MM target: empty^thick 12)MS target: sweaty^wooded 12)MU target: ineffectual^aerodynamic 12)MP target: useless^flowing 12)ML context: Paul hoped he would have a chance to talk to one of the speakers after the conference. He hurried back to the auditorium, but he found it 12)ML target: empty^ideal  MODUNC5.APX ^  242  13)MM context: Elsie, Bob, and Fluffy visited Julie at Christmas. When h( brother phoned a week later to say that he and his dog would be arriving soon, Julie knew she couldn't stand another 13)MM target: assault^profile 13)MP target: guest^stick 13)MS target: animal^career 13)MU target: caller ^pillar 13)ML context: When my father talks about his experiences in the war, I feel lucky that he's still alive. He often tells how just six soldiers were left to defend a whole village against an enemy 13)ML target: assault^collage 14)MM context: Carmen had always tried to think for herself. After she got too drunk at an office party, she wondered how she had been able to follow the 14)MM target: wave^salt 14)MP target: crowd^grass 14)MS target: tune ^doll 14)MU target: horde^blush 14)ML context: The dangers of an earthquake are even greater for people living along a coastline who also have to worry about the possibility of a tidal 14)ML target: wave ^bond 15)MM context: Edmund had written two novels before he turned 30. Later he wrote three more, but his editors rejected them all. They complained that his work had become 15)MM target: stagnant^tapestry 15)MP target: dull^sand 15)MS target: childish^tropical 15)MU target: hackneyed^laminated 15)ML context: It was shocking for Arnie to revisit his old home town after 20 years away. Ugly townhouses replaced open fields, and the pond where he often swam was now 15)ML target: stagnant ^perverse 16)MM context: It was time to call a strike. None of the changes that management proposed had been made, and the workers were starting to make 16)MM target: noises barley 16)MP target: complaints meteorites 16)MS target: mistakes ^arteries 16)MU target: innuendoes^tombstones 16)ML context: The mechanics at Mike's Garage amused one another by imitating the sounds customers would make as they tried to describe their car's odd 16)ML targets: noises^storms  MODUNC5.APX  ^  243  17)MM context: After the disastrous 1929 stock market crash, a lot of  formerly prosperous financiers saw their future prospects suddenly turn 17)MM target: black^north 17)MP target: hopeless^solitary 17)MS target: illegal^ceramic 17)MU target: glum^jade 17)ML context: Fashion is fickle. At different times, both sophisticated society and rebellious youth have focused their tastes around the color 17)ML target: black^sound 18)MM context: The studio audiences seemed to love the new game show host 18)MM 18)MP 18)MS 18)MU 18)ML 18)ML  but the critics and a number of people who watched the show felt that he was too target: plastic^crowded target: artificial^segregated target: short^level target: suave^milky context: One of the biggest problems faced by current environmentally conscious governments is what can be done to stop the production of throwaway target: plastic^gardens  19)MM context: Alice's playmates thought it was funny to pull the  beetle's legs off one by one and watch it struggle to walk, but Alice thought it was 19)MM target: sick^rear 19)MP target: horrible^centered 19)MS target: inspiring^distilled 19)MU target: macabre^rumpled 19)ML context: It is often a difficult challenge for working couples to rearrange their day in order to get adequate home care when their child is 19)ML target: sick^huge 20)MM context: There has been a lot of inefficiency in our administration  of funds. Our new policy is to manage all of our funding under a single 20)MM target: umbrella^traveler 20)MP target: agency^breath 20)MS target: bureaucracy^diffraction 20)MU target: federation^crawlspace 20)ML context: It was a grey, rainy day and Harriet dozed on the bus until she reached her stop. After she had gotten off, she realized that she had forgotten her 20)ML target: umbrella^revision  ,  244  LEXICAL DECISION - SENLEX Odie Geiger - June, 1992 Name:^ Major:^ Class credit to (instructor)?: ^ Is English your most familiar language? Please answer the following questions with 'T' if the meaning of the question is the same as one of the items you were given to read, or "F' if the meaning is different. Some of the items in the experiment were presented more than once with different endings; as long as the question is true of ONE of those you remember, respond 'T'. None of these are trick questions, for example, names were not switched between items, and incidents have not been combined. 1. ^ Alice's playmates tried to get beetles to fight over bits of grain. 2. ^ The politician's presentation was smooth, but his facts seemed suspicious. 3. ^ The new game show host impressed the critics as highly talented. 4.^ It was felt that funds could best be administered under a single agency. 5.^ Tim's teacher's thought that he was manipulative and shallow. 6.^ South America is a cyclist's dream. 7. ^ Edmund's newest novels were masterpieces when compared to his earlier work. 8. ^ All of Einstein's ideas were spectacularly successful. 9.^ Corrine was hungry after a busy morning of classes. 10.^ Jerry's friends found him in a bar, celebrating how well his speech was received. 11. ^ The worker's dissatisfaction with management meant that it was time to call a strike. 12. ^ Muriel and Howard agreed about nearly everything except politics.  ^ 245  ^13.  ^ Bianca saw a great deal of similarity between her life and her mother's life.  14. ^ Don was interested in the French Nobel Prize winner's chemistry lecture, because he had been working on the same ideas himself. 15. ^ Nigel felt that Professor Roger's theories were not well thought out. 16. ^ The lawyers were able to make a clear case against the drug dealers. 17.^ Kevin impressed Elizabeth with his knowledge of cars and baseball. 18.^ Steve tried, but failed, to establish a good relationship with his mother-in-law. 19. ^ Toby gets completely involved in mystery novels by her favourite author. 20. ^ The committee decided to wait before evaluating Laura's performance because she was on a maternity leave. 21. ^ Jack was nervous before his audition for the producer. 22. ^ In the United States each immigrant group tends to maintain its own culture. 23. ^ Andrea found it frustrating to try to teach students who lives revolved around personal concerns. 24. ^ The Department of Education was determined to stand firm on their plans for funding cutbacks. 25. ^ The coach thought he could improve his team's performance by adding some new players. 26. ^ Margie is a good friend, but she tends to be moody. 27. ^ Warren knew that it would be impossible to find a simple solution to all his problems. 28. ^ Julie never tired of entertaining a steady stream of visitors and their pets.  ^  ^29.  ^ Tina felt that drama was only worth seeing if it carried some meaningful message.  30. ^ Pam told Helen that she had planned to elope with Max, but her parents talked them out of it. 31. ^ The recycling committee had trouble finding buyers for the newspapers they collected. 32. ^ Carmen made the mistake of getting drunk and acting like the others instead of thinking for herself.  246  Metaphor Comprehension - 247  APPENDIX F  248  LEXICAL DECISION - DEGRADED CONTEXT - SENLEX9 December, 1992 This experiment consists of a series of short sentence contexts in which the final word is missing. Your task will be to choose which of two words completes the intended meaning of the sentence, and then to indicate your choice by pressing a button corresponding to the chosen word as quickly as possible. The sentence contexts will be displayed all at once on a computer monitor. You are to read the sentences, and once you have understood their meaning, press the bottom buttons on the box as indicated. This will remove the context and present you with two words, side by side, on the video screen. Read them both, and then press either the right or left button on the top of the box corresponding to the position of the word. Once you have made your response, a new sentence context will automatically appear on the screen. As the presentation continues, you will notice that some of the contexts are repeated. However, the endings for these sentences will be different. It is important to the success of this experiment that you read through every context each time it appears, even if you recall it from a previous display. There will be a short pencil and paper quiz following the experiment that you will find easy to answer as long as each context is understood in terms of its endings. If you are ever unfamiliar with the meaning of one or both of the words presented, or if you are unable to decide between the two, or if neither of the words seems to make sense to you in terms of the context, just make a guess and continue on with the experiment. You should try to work at a steady pace, and to be as accurate as possible in making your choice. The entire experiment consists of 200 trials of sentence contexts followed by the quiz, and should take about one hour to complete. You will receive one hour of class credit for your participation. Thank you for your interest. Odie Geiger, Research Assistant Dr. Lawrence Ward, Supervising Professor  249  DEADUNC6.APX EXPERIMENT 6 - DEAD METAPHORS  Note: In this appendix, correct targets are listed first, followed by matched incorrect targets. In the actual experiment, the correct stimulu was randomly selected to appear on either the right or left side of the screen. 1) DD 1)DD 1) DP 1) DS 1)DU 1)DL 1)DL  context: Finding a^ buyer was an unexpected target: bottleneck^paraphrase atmosphere target: difficulty ^welcome target: expense ^ hypodermic target: impediment context: They were^ made so cream would collect in the sketchbook target: bottleneck  2)DD 2) DD 2) DP 2)DS 2)DU 2) DL 2)DL  context: He was personable and very ^ target: bright ^ battle target: smart ^ valid gain target: rich target: astute malted context: The sky was clear and target: bright weight  3)DD 3)DD 3)DP 3)DS 3)DU 3)DL 3)DL  context: His mother-in-law remained target: cold^move ^ seventeenth target: indifferent target: shy^par target: aloof^plaid context: She knew Norway would be target: cold^sort  4)DD 4)DD 4)DP 4)DS 4) DU 4)DL 4)DL  context: They had ^ opposing elections target: positions ^supplies target: opinions ^ footnotes target: loyalties ^oatmeal target: stances context: She felt^ competent when she assumed the steps and employees target: positions  5)DD 5)DD 5)DP 5)DS 5)DU 5)DL 5)DL  context: All he talked about was cars. She had never met anyone so target: shallow^pasture target: superficial^microscopic target: masculine^scratched target: insipid^twirled context: The water is warm and target: shallow^amended  6)DD 6)DD 6)DP 6)DS 6)DU 6)DL 6)DL  context: He spent years searching for a simple tension target: formula target: solution ^ railroad target: faith^ staff snout target: tenet context: Her discovery removed stains. They offered her money for her ^ housing target: formula  DEADUNC6.APX  ^  Friday, June 4, 1993  ^  7) DD 7) DD 7) DP 7)DS 7) DU 7)DL 7) DL  context: I can turn to him for looking target: support regions target: comfort death target: money casualties target: sustenance context: They make sure superstructures have adequate picture target: support  8)DD 8)DD 8)DP 8)DS 8)DU 8)DL 8)DL  context: They maintained their plans were target: firm^cars target: definite^swimming target: progressive ^respiratory target: unequivocal ^chauffeured context: They have textures ranging from soft to target: firm^pool  9)DD 9)DD 9)DP 9)DS 9)DU 9)DL 9)DL  context: She hated entertainment that was too ^unit target: deep ^ released target: profound ^ target: expensive ^ separated deflated target: abstruse context: The^ well was almost 50 meters jazz target: deep  250  10)DD 10)DD 10)DP 10)DS 10)DU 10)DL 10)DL  context: He found the lecture difficult to target: grasp^tumor target: understand^population target: translate^dimension target: comprehend^accelerate context: Her installed handrails so she would have something to target: grasp^quote  11)DD 11) DD 11)DP 11)DS 11) DU 11) DL 11)DL  context: She finds novels so target: absorbing^shuffling target: interesting^responsible target: difficult^committee target: engrossing^unsinkable context: They are developing a material that is highly oil target: absorbing^threading  12)DD 12)DD 12)DP 12)DS 12)DU 12)DL 12)DL  context: They are terrified of riding where roads have inadequate target: shoulders^decisions target: edges^knees target: restaurants^metaphysics target: verges^rivals context: She never went out without a shawl over her target: shoulders^elections  DEADUNC6.APX 13)DD 13)DD 13)DP 13)DS 13)DU 13)DL 13)DL  context: When Pam eloped, Helen was target: shocked^blowing target: surprised^organized glass target: happy prolific target: appalled context: If you use an appliance in the bathtub you will get broader target: shocked  14)DD 14)DD 14)DP 14)DS 14)DU 14)DL 14)DL  context: They had never seen him so rendering target: depressed stained target: unhappy magic target: drunk transluscent target: disheartened context: In the summer the water level becomes dangerously respected target: depressed  15)DD 15)DD 15)DP 15)DS 15)DU 15)DL 15)DL  context: He worried because the smell was so ^ beyond target: strong ^ concert target: intense ^ target: accepted ^ opposite tubular target: pungent context: He exercised to keep himself ^ inside target: strong  16)DD 16)DD 16)DP 16)DS 16)DU 16)DL 16)DL  context: Her life and her mother's were regional target: parallel earlier target: similar porous target: cursed surf target: akin context: It must have been difficult to keep the tracks enormous target: parallel  17)DD 17)DD 17)DP 17)DS 17)DU 17)DL 17)DL  context: The facts were target: clear^north target: obvious^musical target: unusual^highest target: indisputable^aeronautical context: The water is warm and target: clear^alone  18)DD 18)DD 18)DP 18)DS 18)DU 18)DL 18)DL  context: Her feet began to pat target: fly cut target: run slam target: itch rib target: jog context: The robin learned to bid target: fly  251  DEADUNC6.APX 19)DD 19)DD 19)DP 19)DS 19)DU 19)DL 19)DL  ^  context: His reasoning seemed so target: weak^hurt target: vague^brave target: silly^lined target: imprecise^sleepless context: The hostages were extremely target: weak^bare  20)DD context: When he married her, she had no idea he was so 20)DD target: calculating ^fragmentary 20)DP target: shrewd^barred 20)DS target: poor^grey 20)DU target: wily^dank 20)DL context: His balances came out wrong after hours of 20)DL target: calculating ^adolescents  252  MODUNC6.APX  ^  253  EXPERIMENT 6 - METAPHORS Note: In this appendix, correct targets are listed first, followed by matched incorrect targets. In the actual experiment, the correct stimulu was randomly selected to appear on either the right or left side of the screen. 1)MM 1)MM 1)MP 1) MS 1)MU 1) ML 1)ML  context: They decided to wait because she was still target: green^seven target: inexperienced^environmental target: recovering^synonomous target: unversed^embossed context: She thought he would notice but to him it was still target: green^teeth  2)MM 2)MM 2)MP 2)MS 2)MU 2)ML 2)ML  context: Making an organization run requires elaborate target: choreography^astrophysics target: planning^religion target: rules^youth target: designing^figurines context: They were thrilled with such innovative target: choreography^bewilderment  3)MM 3)MM 3)MP 3)MS 3)MU 3)ML 3)ML  context: By noon, she needed target: fuel sums target: food hair target: exercise machines target: sustenance highlights context: He kept a record of his car's consumption of target: fuel cats  4)MM 4)MM 4)MP 4)MS 4)MU 4)ML 4)ML  context: He wanted to do well, but he was target: wobbly clawed target: nervous biggest target: hiccuping garlanded target: flustered vexatious context: Her ladder was target: wobbly ^pitied  5)MM 5)MM 5)MP 5)MS 5)MU 5)ML 5)ML  context: In the U.S., cultures have become ^ target: interwoven ^ morticians target: combined ^ hydrogen target: rebellious ^ bequeathed target: conjoined subtitled context: Shebought fa bric in which strands had been target: interwoven ^reconvened  6)MM context: Not even Einstein's ideas were all 6)MM target: gems pegs ^6)MP target: valuable changing 6)MS target: original returned 6)MU target: laudable buttoned 6)ML context: They can buy artificial materials, believing they are 6)ML target: gems hubs  MODUNC6.APX^ 7)MM 7)MM 7)MP 7)MS 7)MU 7)ML 7)ML  context: She told him it was too much to dilute target: juggle bundle target: manage breathe target: conceal firecracker target: superintend diversion he needed when he learned to context: He found the staple target: juggle  8)MM context: The project was becoming a veto 8)MM target: bear 8)MP target: burden lawyer leash 8)MS target: farce requisition 8)MU target: tribulation 8)ML context: They wanted to photograph a 8)ML target: bear rust 9)MM 9)MM 9)MP 9)MS 9)MU 9)ML 9)ML  context: Her nature is always gentle and target: sunny blade deceased target: cheerful higher target: simple target: genial dotted context: The skies became target: sunny aloud  10)MM 10)MM 10) MP 10)MS 10)MU 10)ML 10)ML  context: She suspected the data had been target: cooked^caring target: changed^talking target: stolen^cosmic target: contrived^nocturnal context: She didn't stop until they were peeled and target: cooked^reared  11)MM 11)MM 11)MP 11) MS 11)MU 11)ML 11)ML  context: His team needed some new trees target: blood samples target: players target: uniforms feathers excerpts target: athletes context: He had to be insensitive to the sight of target: blood truth  12)MM 12)MM 12)MS 12)MU 12)MP 12)ML 12)ML  context: Each^ night she went home feeling target: empty thick wooded target: sweaty ^ ^ aerodynamic target: ineffectual ^ target: useless flowing context: He hurried to the auditorium but found it ^ target: empty ideal  254  MODUNC6.APX^ 13)MM 13)MM 13)MP 13)MS 13)MU 13)ML 13)ML  context: He and his dog were arriving. She couldn't stand another target: assault^profile target: guest^stick target: animal^career target: caller^pillar context: They were left to defend against an enemy target: assault^collage  14)MM context: She wondered how she had been able to follow the 14)MM target: wave^salt 14)MP target: crowd^grass 14)MS target: tune^doll 14)MU target: horde^blush 14)ML context: They worry about a tidal 14)ML target: wave^bond 15)MM 15)MM 15)MP 15)MS 15)MU 15)ML 15)ML  255  context: They complained his work had become target: stagnant^tapestry target: dull^sand target: childish^tropical target: hackneyed^laminated context: The pond where he swam was now target: stagnant^perverse  16)MM context: It was time to strike. The workers were making 16)MM target: noises^barley 16)MP target: complaints^meteorites 16)MS target: mistakes^arteries 16)MU target: innuendoes^tombstones 16)ML context: They tried to describe their car's odd 16)ML targets: noises^storms 17)MM context: They saw their prospects turn 17)MM target: black^north 17)MP target: hopeless^solitary 17)MS target: illegal^ceramic 17)MU target: glum^jade 17)ML context: Both have focused their tastes around 17)ML target: black^sound 18)MM context: Critics felt he was too 18)MM target: plastic^crowded 18)MP target: artificial^segregated 18)MS target: short^level 18)MU target: suave^milky 18)ML context: The problem is to stop the production of throwaway 18)ML target: plastic^gardens  MODUNC6.APX  ^  19)MM context: They pulled its legs off. She thought it was 19)MM target: sick^rear 19)MP target: horrible^centered 19)MS target: inspiring^distilled 19)MU target: macabre^rumpled 19)ML context: It is difficult to get care when a child is  19)ML target: sick^huge  20)MM context: Our policy is to manage our funding under a single 20)MM target: umbrella^traveler 20)MP target: agency^breath 20)MS target: bureaucracy^diffraction 20)MU target: federation^crawlspace 20)ML context: She had forgotten her 20)ML target: umbrella^revision  256  ^  257 LEXICAL DECISION - DEGRADED CONTEXT - SENLEX9 December, 1992 Name: ^ Major: ^ Class credit to (instructor)?: ^ Is English your most familiar language?: ^ Please answer the following questions with "T" if the meaning of the question is the same as one of the items you were given to read, or "F" if the meaning is different. Some of the items in the experiment were presented more than once with different endings; as long as the question is true of ONE of those you remember, respond "T". 1.  She watched it walk briskly away.  2.  She was suspicious of some of the data.  3.  He impressed the critics with his talent.  4.  It was felt that funds could best be administered through a single agency.  5.  He was appeared to be personable, but he was manipulative.  6.  They enjoy riding on narrow, busy roads.  7. ^ They complained his work was too minimalist and literary. 8.^All of Einstein's ideas were spectacularly successful. 9. ^ She was hungry after a busy morning. 10.^His friends joined him in celebrating his success. 11. ^ The worker's frustrations made it necessary for them to call a strike. 12.^Each of them had opposing views. 13. ^ She saw a great deal of similarity between her life and her mother's. 14. ^ He found the lecture extremely interesting and informative. 15. ^ He thought the theories were not well thought out. 16. ^ Thee facts of the case were straight-forward. 17. ^ She was impressed by his knowledge of cars.  ^ 258  ^18. ^ He tried, but failed, to establish a good relationship with his mother-in-law. 19. ^ She gets completely involved in the novels she reads. 20. ^ They decided to wait before evaluating her performance because she was on maternity leave. 21.^He wanted to do well, but he was nervous. 22. ^ In the U.S., cultures have tried to maintain their own identity. 23. ^ She went home frustrated and exhausted each night. 24. ^ They were determined to stand firm on their original plans. 25. ^ The team would benefit by getting some new players. 26. ^ She is usually moody, and often mean. 27. ^ He found the answer to all his problems in a fortune cookie. 28. ^ She never tired of entertaining her friends and their pets. 29. ^ She felt drama was only worth seeing if it carried a meaningful message. 30. ^ When Pam wanted to elope, Helen talked her out of it. 31. ^ They had trouble finding a buyer. 32. ^ She wondered how she could have gone along with the others, instead of thinking for herself.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0098794/manifest

Comment

Related Items