UNIT Th THE COGNITIVE DIMENSION OF MEANINGLesson 9 AN INTRODUCTION TO COGNITIVE SEMANTICSIntroduction “9.1. The relationship between lingui 9.2. Approaches 10 uistic knowledge and cognition, 92.2. ce Soe te Different views. 9.2.3, The nature of concepts. 9.24. TI 9.2.5, The standard 9.3, The mental approach. ‘approach. ferent approaches are different approaches 0 the ies to the issue of categors ee ization. question of the relationship INTRODUCTION tn lesson 2 we studied the connection between linguistic models and the different views that semanticists hold, depending on their previous tices about how the external world is perceived and codified by speakers, een oe earned that there are some important concepts that crop uP Aga i again in semantics; we called them conceptual tools and one of them ‘was the concept of categorization. in this lesson we will learn some more about the different ways ti Which human beings try to categorize and thus understand the world ochd them and how allthis affects the way we understand meanl ni We call this approach to the understanding of meaning cognitive because it has deep roots in cognitive psychology. ‘We will also study the nature of concepis and the two main approaches to conceptualization, and we wil finally link the different approaches Tinguistic analysis and to conceptualization models. In this section we will ‘be following Cruse 2004 and Saeed 2003 quite closely. 9.4, THE RELATIONSHIP. BETWEEN LINGUISTIC KNOWLEDGE AND COGNITION One of the most widely accepted views among cognitive linguists is the nda that that there is no separation of linguistic knowledge from general thinking. In this sense they strongly oppose the influential views Pf other linguists, such as Chomsky and Fodor, who see linguistic behaviour seein ether separate part of the general cognitive abilities which allow Jearning and reasoning. Formal and functional approaches to grammar are usually linked to certain views of language and cognition. For instance, generative graninat fs generally associated with the idea that knowledge of linguistic structuresSEMANTICS and rules forms ay in autonomous modi fan omous module or faculty arate ene Mendon Tera Pa usepeen = module is often combined with a view of iPaslons, , and semantics, s ee, tally Suotoctod the ee ing: Ss independent modules. This view ae re i studied without reference to semantic conta Principles can be Functionalist approaches, whi nekunetionalie es, which, as theo with which cognitive ident lipases rears: oe eri imply a diferent view of language altogether oe explains that tela Uncen dae eee and he adds that oe polaris hey tight roto esr ad that it makes sense to look for principles shared a ‘ange of copnve domains. Similar he argues that no adequate account of rammati Possible without taking the meaning of elements Bache ot the most interesting characte itiate between linguistic ledge encyclo real world meet A eho knowledge. From an extreme point of view the glances eae -al patterns cannot be given in te shee es ies but only in terms of the speaker inte contexts of language use. peat tics of cognitive linguistics is ae 7 gain, defines as the ‘doctrine’ of objective reference there is an object herr is an objectively correct way to associate symbols with things in th e One alternative a Proposal withii ie called experientialis ‘thin the cognitive linguisti. J have meaning ny Pee ain iat swords and ieee ie Miiasted ene aes of our interaction with the world. roeral 1. Meani ee not stem from an abstract and fi ing is symbols and things ir ct and fixed correspondei beings interact wi gs in the world but fi ppengleace ‘ith the world. rom the way we hur dynamic patter rid. We human beings h: i ee ns of interaction with gs have certain recurrin, orientation, manipulation with a physical world th ing am of object hrough spatial Patterns structure and constrain how ive commey ge eeamnmning. These meaning, i AN INTRODUCTION 10 COGNEFIVEIBEMANT > ‘Embodiment as proposed by Johnson, 1987, Lakoff, 1987, and Lakoff and johnson, 1999, constitutes w central element in the cognitive paradigm, Th this sense our conceptual and linguistic system and its respective categories are constrained by the ways in which we, as huni beings, perceive, categorize and symbolize experience, Linguistic codification is Prtimately grounded in experience: bodily, physical, social and cultural, 9,2. APPROACHES TO CATEGORIZATION. DIFFERENT VIEWS 9.2.2. Concepts seen as an essential survival element that Conceptualization can be born totally defenceless, it would characterizes the human species. Being Prat'been very difficult for the human offspring to survive, if it were Ail far his powerful “understanding of the real world’ features Conceptualization that language allows has been essential for us survive co a beminate other less intellectually endowed species. If someone heath ‘Beware of snakes in the trail, itis the understanding of the concept [SNAKE] aenatne pond snake triggers that allows the hearer to be aware of potential danger It's this abstraction potential of concepts that helps Ws 19 navigate ae Mase chaoti surrounding word. Because speaker and hearer Sale t category [SNAKE], communication between them has been possible, Cruse explains how concepts are vital to the efficent functioning of Juma cognition, and he defines them as organized bundles of stored ramvvizdge which represent an articulation of events, entities, situation ace ur experience. If we were not able to assign aspects of Out ‘experience to stable categories, the world around us would remain Guorganized chaos. We would not be able to learn because each fexperience would be unique. It is only because we can put simi (but exe tical elements of experience into categories that we can recounie aoa de having happened before, and we can access stored knowledg iitnt them, Shared categories can be seen then as a prerequisite (0 communication. Cruse (2004) proposes what he cals a fairly simplistic model both of the structure of the conceptual system and of the relations betwen Linguistic forms and concepts. Concepts are linked together in complex i network as follows. The links are of specific types (eg, is a kind of, isa part of, is used for, lives in, etc.) and are of variable strength. i10 SEMANTICS NT] coe horse hoot These links corres py orrespond to concepts of a mi i fhe concepts which they serve to connect, hich ae gaia ricer sist coanerctertes Linguistic forms map onto conceptual smaeied te wangerable complexity although Cruse confines his atten india orgs only and states that the word horse, for example, has a diet Ik cept [HORSE] only and not to the concem [atanat a 7 AN INTRODUCTION TO! 191 0.2.4, THE CLASSICAL APPROACH 9.2.4.1. Problems of the classical approach In lesson 2 we studied how the classical view of categorization describes word meaning as a set of criterial properties or features. According to this theory, categories have clear boundaries, as membership is limited to those entities possessing the conjunction of necessary and sufficient features particular to the category in question. Within the category itself all prembers have equal status thus, the main characteristic of the classical theory of categorization is that is has fixed, well delimited boundaries. However there are important limitations and problems to this approach. In Witigenstein’s famous example of the concept of games He argued that is was impossible to draw up a list of features possessed all games which jointly distinguish games from non-games. Following his example, if we suggest the following list of features below Wt characterizing the concept of games , ‘ a. involves winning and losing r b. involves more than one person c. has arbitrary rules d. done purely for enjoyment ‘we will see that it is not possible to characterize all games since there are ‘activities that we call games which do not satisfy these features. However, in spite of lack of compliance with the above criteria, we can communicate using the word game perfectly successfully. ‘A large body of research on category structure demonstrates that the boundaries of natural categories are fuzzy and contextually flexible, For example, Berlin and Kay (1969) studied colour categories from a psycholinguistic and anthropological point of view, and they found that While judgements of central examples of colours were relatively constant across subjects and reliable within subjects on different occasions, judgements of borderline instances of colours, such as between red and Orange, or blue and purple, showed neither agreement among subjects por reliability within subjects on different occasions. In addition, Labov(1973) studied subjects’ naming of line drawing, illustrating cu mugs, vases, bowls etc, that systematically varied parameters such as ratio of height to width, curved or straight sides and presence or absence ‘of handle, and he found that certain items received reliable assignation toa particular category, while others were uncertain. He also found that aaye SEMANTICS contextual conditio, instance, an inet o Alter the subject's responses, so that, for imagine all the items containing rice extended t he boundaries of the BOWL category, while a similar instruction to extended the CUP category, imagine coffee as contents 9.2.5, The standard prototype approach The standard prototype approach derives i research done in the 70s by Rosch et alia (1973), Her en mera i Her main contributi Stctaved es as £0 argue that natural conceptual eateporis ace 3” examples or prototypes i and that other items are assimilated toa catcooy nfs ca egTHeS, they sufficiently resemble the prototype or nat) eas © whether y good example example irly good example joderately good example ily poor example example bad example/ not an example at all. ing to Cruse’s, the following exampl eel iple shows the appli ‘above to the ratings given to the category VEGETABLE on eee eS a POTATO, CARROT TURNIP, CABBAGE CELERY, BEETROOT AUBERGINE,COURGETTE PARSLEY, BASIL RHUBARB LEMON Nlolala}els|— The prototypes of categories are determined by selecting the item with the lowest average numerical score. Ratings of GOE are strongly culturally dependent, What is prototypical fruit in a British context is not the same in a Muslim cultsin, Ina British context, DATE typically receives a GOE score of 35 to the category of FRUIT, but in a group of Jordanians it obtained almost unanimous 1. Family resemblance The philosopher Wittgenstein introduced the concept of resemblance. He explained that the members of a family typically one another, but there may well not be any set of features that they possess, and it may be possible to find two members who have no features: in common. However, they will be linked by a chain of intermediate members with whom they do share features. So, for example, A may have ‘no features in common with C, but has the same nose as B, who in turn has the same eyes as C. Prototype theory includes Wittgenstein’: notion that family resemblance unites the members of a category and includes the important concept of central and peripheral members. Categories thus have internal structure. There are central members, less central members and peripheral members; and there are also borderline cases. 9.2.5.1. Prototype effects. Cruse proposes a list of important effects of prototype theory. He argues that, taken in isolation, the existence of GOE scores may not be particularly i4 SEMANTICS relevant, but prototypicality, as measured scores, a by GOE i with important aspects of cognitive behaviour such as the following: Order of mention When subjects are asked to list m a embers of a category, and ci if they are put under time Pressure, the order of listing correlamme rd GOE ratings, with the prototypical member showing a s ndency ; er showing a strong tenden Overall frequency The overall fre sntion i i GOR CNET frequency of mention in these lists also courelates with Order of acquisition a Cf 7 Prototypical members of categories tend to be acquired first, and orde1 following, cS " Vocabulary: learning Children at later sta; ges of language acquisition, when ° ) When v enlargement can be greatly influenced by explicit teaching learn rae of verification In psycholinguistic experiments in whi experiments in which subjects are asked t quickly as they can in a categorization task, subjects produce fone prototype theory, itis usually held that or 7 nly the protot er cont menership of category the degre of membership ater lent on their d Unum selected by their COL scares ance t@ the Prorye, AN INTRODUCTION we 9.3. THE MENTAL REPRESENTATION OF CATEGORIES Cruse explains that there are recent feature-based treatments of prototype structure where categories with a prototype structure are represented by a set of features. This development shows a combination, Of classical and prototypical approaches to the mental representation of categories, However, unlike the classical features, these do not constitute a set of necessary and sufficient criteria, except for the prototype itself. Rather the features are such that the more of them that are manifest in some particular instantiation, the higher the GEO score the item in question will obtain, In such systems, features are weighed differently depending on how close to the central feature they are. The following example illustrates the description of the category VEHICLE in a not necessarily exhaustive number of features: a, Designed to go on roads b. Has ist own propulsive power cc. Can go faster than an unaided human d. Can carry persons/goods in addition to driver ¢. Has four wheels £, Metallic construction g. Persons/goods enclosed h, Manoeuvrable A central example of the category VEHICLE, such as car will have all ‘those features. However, the following items would have missing elements such as in TRAIN: Not designed to go on roads Not manoeuvrable TRACTOR Not designed to go on roads Driver not always enclosed BICYCLE Does not have its own propulsive power ‘We then see that the category VEHICLE, like GAME is one for which it is not possible draw set of necessary and sufficient features. a= sumantics a" Basie level categories F tod wo wl sady he cheese a. vehicle- car- hatchback b. fruit-apple-Granny Smith ¢. living thing- creature- animal- cat - Manx cat d. object- implement - cutlery — spoon different categories of entities jiven by Cruse. fo ~ teaspoon e basic level or generic level of specificity has various characteristics 1, Ttis the most inclusive level at whi charact ; Cn at which th i of behavioural interaction. For example. you conser oreo animal unless you are told which animal tomimic. 2. Tis the most inclusive level for which a clear visual image can be formed. Again one cannot visuali; ee men one = Visualize an item of cutlery or fruit if 3. Basic level items are used fe si for neutral, everyday refe f ‘often felt by speakers as being the real name ofthe relerer |. The basic level is the level at whi ‘ at which the ‘best’ categori created. And good categories include the following cham, o Distinctness from neighbouring categories : ii, Within-category resemblance fii, Informativeness: the amount I : t of informati if we know that something belongs to that 5. The names of basic-level ca simple and they are not categories. For example, all the rest have more c spoon tec. ion we gain access to category itegories tend ta he morph. i phological; X metaphorical extensions from other if we take spoon as the basic-level term. “omplex names: teaspoon, tablespoon, coffe. Cruse studies a number of se studies problem: categorization Firs, he explains how the basis of COS aan toed ina comenseety-name and on item name, is not enough. The ‘GOE a es nfl lore basic scales such as familiarity and well- ‘Secondly, one of the most serious shortcomings of ‘standard prototype igs of st theory is that no category boundary is recognized, However aren legory Without a boundary is virtually useless because the primary function of no AN INTRODUCTION TO COGNITIVE BEMANTICS 197 a category is to discriminate between things which are in it and things which are not in it, Cruse takes the view that, classical theory of categorization with necessary and sufficient features, sets a boundary but allows internal structure and if one rejects it, rejects both important features. As a result, he proposes that a fully satisfactory description of a category must specify both internal structure and location of boundary area. It is accepted that category boundaries are more or less fuzzy but even fuzzy boundaries have locations. Finally, there are also problems with the degree of membership. For ‘example, it seems that BICYCLE and SKATEBOARD are borderline cases of the category VEHICLE. Thus the notion of degree applies only to such borderline categories. 0.3.2. Types of conceptual category The characteristics of the category NATURAL CONCEPTUAL CATEGORY include distinguishing clearly between things that are in it ‘and things that are not in it. That is they must have well-defined boundaries. In addition, the major function of conceptual categories is to provide headings under which information can be economically stored, 9.4, THE CONCEPT OF FRAMES Concepts cannot be treated in isolation because every concept is embedded in a larger body of knowledge of some sort. Understanding any concept requires taking into account wider domains. This is the main idea underlying Fillmore’s frame semantics. Fillmore put forward this idea as a better alternative to feature theories of word meaning. ‘To understand the concept MENU in the sentence John asked to see the menu, the concept MENU requires a wider contextual framework to be understood. This concept, whose basic meaning is ‘a printed list of food items’, requires some understanding of a scene in a restaurant or café, the idea of customer, waiter, sequence of meals, ete. ‘That is, some words which at first sight seem to be explicable by feature analysis, on closer examination turn out to require appropriate frames to be activated. on1 ae 6 pt A similar notion wi — proposed i. r once nlm sense hen ewe atthe bankas ee Inclusive nature, Crass sat themselves concepts of a more meaelor {nelusive nature, Cruse explains how the concept of FINGER. separate fe ir Neste pe the wheel of a bicycle fa Gea eee ‘one another. If we consider ree wheel ; ence to a wider domain such as bicycle or wheelbarrow Tangacker refers to the region or aspect of a domain highli ted by a t of | lomain et ighligl in this way as the base. : of the hace BICYCL a ont Previous examples, WHEEL profiles a region Profile and base, on the othe . hand : ones For camps FINGER functions are not absolute terms but relational KNUCKLE and NAIL. form more specific proflings, are domains thet ape no pect oF inclusiveness in the sense that thy See raans that are not profiles of anything more inclusive; these are Fi, araraamaits and include elementary notions, such as SPACE, . J ITY, and . ia Jackendoffs basic ontological eategories NO TMS 47° similar to 9.5. FRAMES OR IDEALIZED COGNITIVE MODELS (ICMs) Both Fillmore (1982) and Lakoff i ly that links linguistic pasion te fies iden c peut claim that speakers have folk theories about the Suk. baa -expevience and tooted in thei culture Fillmore calls these theories Lakoff af idealized a cognitive models. (ICMs). According to Saeed, ical marriage ICM or fram arriage ICM c : a monogamous ini gible le, typically involving romantic love, ete Inthigmedel ga ens aply the term to celibate priests, Robinson Cruso@.o1Datzan. In thie is view, ll AN INTRODUCTION 10 COGNEFIVIE SEMANTICS: 199 ‘using a word involves combining semantic knowledge and enclyclopaedic, knowledge, and this interaction may lead to typicality effects. Lakoff (1987: 68) defines ICMs as the way in which we organize our knowledge. ICMs may also be defined as cognitive structures whose purpose is to represent reality from a certain perspective in such a way that they result in a process of idealization of reality. Other authors such as Ungerer and Schmid (1996: 48-49) also provide ‘amore detailed description of the notion of ICM. They describe them with some additional characteristics of cognitive models: they are basically ‘open-ended, they tend to build networks since they are interrelated entities, and they are omnipresent. ICMs use different kinds of structuring. principles (see Lakoff, 1987: 68): propositional structure as in Fillmore’ Frame Semantics, image-schematic structure of the kind deseribed it Langacker’s Cognitive Grammar, and metaphoric and metonymic mappings ‘as described by Lakoff and Johnson (1980, 1999). We will look at some ‘aspects of Langaker’s Cognitive Grammar in lesson 10. SUGGESTED READINGS i — For the relationship between linguistic knowledge and cognition | see Saeed (2001: 299-302) (2003: 342-43). — Fora clear exposition of the main characteristics of categorization, both in the classical account and in the prototype theory, see first Saeed (2003: 32-47) and then Cruse (2000: 130-137; 2004: 1251. 39), ANNOTATED REFERENCES Cuenca, M. J. and Hitrerry, J. 1999. Introduccién a ta lingitstica cognitiva. Barcelona: Ariel Lingitistica. ‘This is an excellent introduction to cognitive linguistics for the Spanish readership. It deals with the origins of cognitive linguistics, categorization, semantic structure, metaphor and metonymy, polysemy and radial categories, grammaticalization, and the present and future of cognitive linguistics. Kxeter, G. 1995, La semdntica de prototipos. Madrid: Visor. Kleiber provides very interesting discussion of issues related to categorization: the standard classical version of categorization, an overall review of the theory of prototypes) the standard version of prototypicality and its problems, et aii200 SEMANTICS Laxorr, G. 1987. Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago: University of Chicago Press. Lakoff’ work is recommended to those who have already read about cognitive linguistics and want to get a more detailed and exhaustive view of some central issues within the cognitive paradigm. Its written by one of the leading proponents of cognitive linguistics, even though the author ‘was initially adscribed to the generative movement. He later tumed to the ‘cognitive paradigm and has written many excellent works on several central topics in cognitive semantics, especially metaphor, Lakorr, G. and Jounson, M. 1980. Metaphors We Live By. Chicago: Chicago University Press. These authors offer a very readable overview of many aspects within the cognitive paradigm, especially metaphor and its classification. In addition, some chapters are devoted to the description of objectivism andl experientialism. Laxorr, G. and JOHNSON, M. 1999. Philosophy in the Flesh. The Embodied Mind and its Challenge to Western Thought. New York: Basic Books, ‘This work deals with many of the topics covered in the 1980 book by the same authors. However, it is much more exhaustive than the othe. phe and includes the developments made within the cognitive paradigm from 1980 to 1999. For instance, the cognitive theory of metaphor ts enriched by Grady’s theory of primary metaphors. ‘Taytor, J. 1995 [1989]. Linguistic Categorization, Prototypes in Linguistic Theory. Oxford: Clarendon Press. j\very lengthy account of the theory of categorization from its origins the appearance of prototype theory is offered here. The central tenets of cognitive linguistics have been applied to semantic ‘This book is an essential introduction to this area of research, has been the object of both positive and negative criticism: according Hasser’ review in Cognitive Linguistics (11/2000), there are important tions to the text such as the author's claim that the cognitive: sed are mentally represented when this issue is still under scrutiny, the other hand, Niemeyer, from the University of Bremen, was a mor ful reviewer in her 1998 review (Cognitive Linguistics 9(3)), The book #s the main topics dealt with in any introductory book to cognitive istics, 201 AN INTRODUCTION TO COGNETIVEE SIMANTICS GENERAL REFERENCES: i tution. sun, B. and Ke, P (1969). Basic Colour Term: Ther Universality and Evolution. i: ey oC yaaa ame semantics. In Linguistics S Nee riers the morsing cake, 11138. ‘Seoul: Hanshin, jell wy, W. 1973, "The boundaries ofthe words and their meanings In 3. Ne ean ty and RLW. ShuyCeds) New ways of analyzing varia Washington DC: Georgetown University Press. ie [Laxor®, G. and M. JOHNSON. 1980. Metaphors we live by. Chicago and London: University of Chicago Press. on Laxorr, G. 1987, Women, Fre and Dangerous Tings, What reveal the mind. Chicago and London: fieses — — Lara, 1988, Coin Smt ne oe Gel Ra feaning and Ment s Po ; Rost, EH. 1973 ‘Naturl categorie’ Corie Pooley 7 _ ‘Tavior, J. 1995 [1989]. Linguistic Categorization. Prototypes ‘Oxford: Clarendon Press. conti UnceR, F.and ScHMID, H.J. 1996. An Introduction to Cog London: Longman. EXERCISES AND ACTIVITIES 1. How can these concepts be better defined? Mark them accordingly Classically Prototypically bieycle love ‘green explanation : ae. 2, Define the following words in terms of binary features: robin, ostriel ” bachelor, spinster: 3, Give a set of prototype features for one or more of the following or Le 't oni va SICAL INSTRUMENT, HOBBY, BUILDING, CLOTHES, HOUSEHOLD202 ‘SEMANTICS —_—_— anes 4. Which of the following would you consider to be basic-level categories? (Cruse, 2004) 5. Explain how illustrationsin lessons 8, 9, 10 and 11 in pages 169, 183, 201, 227 can be better defined, 6. Explain what is odd in illustration 10,