theory to language use in the foreign language classroom is restricted to very few theoretically sound studies. Some of these will be discussed in more detail in section 3. The chapter is organized in the following fashion: Section 2 discusses some of the main tenets of Second Language Acquisition in the light of linguistic theories and relates them to the Cognitive Linguistics enterprise. Section 3 gives a brief overview of studies which so far have dealt with pedagogical considerations in light of the theory of Cognitive Linguistics. Section 4 outlines in detail the major mental principles or operations, such as iconicity, construal, and prototypicality, which are relevant for a didactic application of cognitive linguistic theory to practical fields such as organized language learning. Section 5 presents the main ideas, methodologies, and results of some of these studies in more detail by briefly exploring the teaching and learning strategies of specific grammatical and lexical constructions such as phrasal verbs and phraseology. In conclusion, section 6 offers an outlook on future research. 2. Cognitive Linguistics and Second Language Acquisition Before outlining a Cognitive Linguistics inspired approach to language pedagogy, I will briefly describe and evaluate, from a cognitive linguistic perspective, earlier accounts of foreign-language instruction and methods of grammar teaching which were formulated in the wake of successive approaches to linguistic theory. Historically, it has always been a hotly debated question as to whether gram- mar should be taught deductively or inductively (Johnson and Johnson 1998: 146– 48). On the deductive approach, learners are supposed to consciously learn the rules of grammar, and they should possess an explicit, metalinguistic knowledge of these rules. The rationale is based on the traditional approach to the teaching of Latin and is commonly described by its detractors as the grammar-translation method. Sets of grammatical rules and long lists of words have to be memorized, and the written language rather than the spoken language is emphasized. Cogni- tive Linguistics, likewise, offers the learner so-called rules for correct usage, but the Cognitive Grammar conception of a grammatical rule takes the form of a con- structional schema, a generalization over a set of linguistic expressions (Achard 1997: 164). Furthermore, Cognitive Linguistics does not focus on the violation of some arbitrary rule of syntax; rather, it assumes that syntactic structures are subject to a semantic explanation, as forms which symbolize meanings (Taylor 1993). The inductive approach, however, argues that the rules of foreign languages may be induced by learners if language input is organized and offered in a system- atic way. This view developed as a result of the structural approach to linguistics, which was geared toward analyzing human language in terms of minimally con- trasting units. The structural syllabus was mainly associated with the method of audiolingualism in the 1950s, which focused on sentence patterns as the unit of 1140 martin pu ¨ tz analysis. In the behaviorist climate of Bloomfieldian Structuralism, the discovery of sentence patterns was coupled with pattern drills. In the wake of Cognitive Lin- guistics, it has become clear that patterns (now called ‘‘constructions’’) are linguistic realities indeed governing a large amount of language use. This does not mean, however, that the process of sentence construction is largely determined by the grammatical properties of linguistic units, such as words—as has traditionally been held in formal linguistics (e.g., Harris 1964). From a cognitive linguistic perspective, meaning, rather than grammar, is unarguably the primary determinant of whether linguistic units can combine with each other (Lee 2001: 70). The audiolingual approach was soon called into question in the 1960s follow- ing Chomsky’s (1959) devastating criticism of Skinner’s Behaviorism and his view of learning. Chomsky proposed a mentalist approach to acquisition, whereby sen- tences are not learned by stimulus-response drills on patterns but generated from the learner’s underlying competence. Chomsky’s generative theory of grammar and language acquisition involving an autonomous Language Acquisition Device only led to a didactic oversimplification, equating first and second language, with no rule formulation and mainly rich learning environments for the Language Acquisi- tion Device to operate. 1 Chomsky’s narrowing down of linguistic competence to grammatical competence and its intuitive strategies of acquisition provoked a re- action from functionalist and sociolinguistic approaches to language, synthesized by Hymes (1974) in terms of the notion ‘‘communicative competence’’ and in terms of his ‘‘ethnography of communication.’’ Following the rise of sociolinguistics and of functionalism in theoretical linguistics (Halliday 1985), a great deal of attention was indeed being paid to the social and functional aspects of language use. In the area of language teaching, this, in turn, led to the ‘‘communicative approach’’—a largely British innovation. It opposes the view that consciously learning the gram- mar of a language will result in an ability to use that language in social interaction. More specifically, its focus is on a functional account of language use which places emphasis on language as an instrument for conveying meanings in social situations. In other words, the functional and communicative potential of language should be emphasized (e.g., ‘requests’, ‘denials’, ‘offers’, ‘complaints’, etc.), rather than the mere mastery of formal structures (e.g., phonological, grammatical). Communi- cative competence, the ability to use the linguistic system effectively and appro- priately, is the desired goal. These characteristics of the communicative view of language are in line with Richards and Rodgers’s (1986: 71) assumptions that (i) the primary function of language is for interaction and communication and (ii) the structure of language reflects its functional and communicative uses. This last point raises the need to bridge the gap between, on the one hand, the structural aspects of Second Language Acquisition and, on the other hand, its func- tional and sociolinguistic aspects. In other words, a linguistic theory is needed which stresses the conceptual link between the form and the function of language. Clearly, nothing is to be expected here from Generative Grammar, where Second Language Acquisition research has been mainly syntactic in nature, abstracted from social and functional considerations. Achard (1997) illustrates the issue by means of a syntactic analysis of English modals: the generative paradigm simply claims the existence of an cognitive linguistics and applied linguistics 1141 innate universal grammar providing the possible parameters for language and uses a parameter-setting approach depending on which specific language is involved; it views acquisition as complete ‘‘once the appropriate parameters have been set pro- perly’ ’ ( 162). Given the belief in an autonomous language module, functional con- siderations of the conventional and social usage of modals by the young child are not taken into account. Generativists adhere to the theorem that the use or function of language is something analytically distinct from the structure of language. As such, a model is needed which emphasizes that language relates to our conceptual world and our human experience in such a way that every grammatical construction reflects its conceptual experiential value. It is precisely the theoretical model of Cognitive Grammar which represents a valid framework within which Second Language Acquisition research may take place, especially ‘‘because it af- fords a satisfying conceptual integration of the structural and social aspects of L2 acquisition’’ (Achard 1997: 159). Apprehension of the physical, social, cultural, and linguistic context is implicitly acknowledged in Langacker’s dynamic usage-based model (see section 4.2 in more detail), which focuses on the actual use of the lin- guistic system and a speaker’s knowledge of this use. Basically, the model claims that linguistic units are abstracted from usage events, that is, the actual instances of language use, and that such events consist of ‘‘a comprehensive conceptualization, comprising an expression’s full contextual understanding, paired with an elaborate vocalization, in all its phonetic detail’’ (Langacker 2001b: 144). The contextual facets of Cognitive Linguistics including the social, cultural, and discourse ingredients of language can therefore be exploited for a communicative and usage-based ap- proach to language teaching in the classroom. Independently of Cognitive Linguistics, another trend opposing Chomsky’s views developed, the so-called cognitive-code learning theory, which allows for a conscious focus on grammar and a recognition of the role of abstract mental pro- cessing in language learning. This view implies that learners should be made aware of the correspondences between varying structures and that grammar can be taught and learned deductively (Johnson and Johnson 1998: 149). Although there has been considerable interest in the implication of the cognitive-code theory for language teaching, no particular method incorporating this view of learning has emerged (Richards and Rodgers 1986: 60). 3. Overview of Applied Cognitive Linguistics Oriented Studies Having positioned Cognitive Linguistics within the context of a number of models and methods of Second Language Acquisition, I will now present a selection of studies which are viewed as instances of the cognitive linguistic approach to language 1142 martin pu ¨ tz pedagogy. Given the duality in early Cognitive Linguistics between Langacker’s concentration on grammar and Lakoff’s conception of the world of thought via metaphor research, we can expect these two trends to emerge in Applied Cognitive Linguistics, too. One of the first linguists to discuss in detail the cognitive-didactic approach to English grammar is Dirven (1989a), who investigated where Cognitive Linguistics can make a contribution to the general process of facilitating language learning. More specifically, the following four major tenets of Cognitive Linguistics were dealt with in the light of one learning problem, namely that of the English modality sys- tem: (i) the unity of linguistic levels (i.e., morphology, syntax, and semantics); (ii) the role of context for a linguistic expression; (iii) the concepts of profile and base in the characterization of linguistic expressions; and (iv) the concepts of prototypes and schematicity, the former covering the more frequent senses of expressions and the latter representing the commonality between all senses. The analysis shows that in general a pedagogical grammar of English is bound to analyze language-specific forms from a categorizing perspective, that is, to uncover the conceptualizations encoded in linguistic expressions. With respect to the issue of Foreign Language Learning, this means that cross-linguistic contrasts between conceptualizations must be identified in order to facilitate the learning process (see section 4.5). The first systematic and principled account of the application of cognitive linguistic insights in the teaching and learning of grammar is provided by Taylor (1993). He starts from the process of ‘‘consciousness raising,’’ which, in the wake of the contrastive linguistics approach of the 1960s and 1970s, had been developed as a new insight in foreign language pedagogy, particularly as a counterbalance against purely intuitive communicative learning. Given its view of meaning as largely iden- tical to conceptualization, Cognitive Linguistics can only strongly support the role of consciousness in language learning, at the same time emphasizing other cognitive linguistic principles such as the notion of imagery and rules in Cognitive Grammar. Central to Taylor’s account is the general cognitive assumption that syntax is mo- tivated by semantics and that therefore the perceived arbitrariness of the foreign language system must be reduced and its motivated structures explained to the language learner. In a less didactic contribution, but intended as a fragment of a pedagogical grammar of English, Dirven and Taylor (1994) are concerned with the basic conceptualizations of modal auxiliaries such as can, may, and must/have to.On the basis of the schematic meanings of modal auxiliaries, the different domains of modality (e.g., potentiality, necessity, desirability) and the forms that are used in each domain, for example can/could (ability), may (permission), may have/might have (potentiality), are identified. Although the didactic aspect is not explicitly out- lined in this paper, we may conclude from the analysis that a schematic account of the basic meanings of the modal auxiliaries provides cognitive insight in the rule complexes of English modality, thus facilitating the language learning process. Given the immediate impact of Lakoff and Johnson’s Metaphors We Live By (1980), it is no wonder that by far the bulk of applied cognitive linguistic studies concentrates on the issue of metaphors and, as a corollary, on the area of vocabulary cognitive linguistics and applied linguistics 1143 acquisition without, however, exploiting the didactic and methodological tools of Cognitive Linguistics in a systematic way. The results from the numerous studies on metaphor can obviously not be ignored and will have to find their way into new strategies and methods of language teaching and learning. Most studies on metaphor provide predominantly theoretical accounts of the cognitive underpinnings to language teaching (e.g., Low 1988; MacLennan 1994; Radden 1994; Barcelona 2001) or engage in experimental studies to demonstrate the usefulness of the cognitive approach in foreign language pedagogy (e.g., Lazar 1996; Deignan, Gabrys ´ , and Solska 1997; Verspoor 1997; Boers and Demecheleer 1998; Boers 2000a, 2000b). Methodological approaches to figurative language and met- aphor are proposed by Lazar (1996) and Lindstromberg (1996), who suggest ap- propriate teaching techniques and sample procedures. One of the first studies to give metaphor a more prominent place in language teaching from a cognitive linguistic perspective is Low (1988), which looks at the functions of metaphor in language use and the pedagogical implications for devising teaching and reference materials. Low argues quite convincingly that the systematicity of metaphor re- quires a discussion of methodological problems, such as constraints on the design of teaching materials and the development of effective types of exercise (e.g., multi- text and multitask activities). A more theoretical account of the implications of cognitive insights to Foreign Language Learning is provided by Radden (1994), who discusses the importance of image schemas and conceptual metaphors in order to make explicit to the learner the systematic coherence of metaphorical expressions in language use. Central to his claim is the idea that a considerable part of the lex- icon is iconically motivated and therefore cognitively easier to grasp for the lan- guage learner (see section 4.1). Boers (2000b), then, introduces Cognitive Seman- tics into the field of ‘‘English for Specific Purposes’’ and explores the potential benefits of an enhanced metaphorical awareness on the part of the language learner. Such an enhanced metaphorical awareness may be achieved by drawing students’ attention to the source domain or the origin of the figurative expressions (for in- stance, in socioeconomic discourse). As the results of a small-scale experiment show, enhanced metaphorical awareness may indeed help learners to better retain unfamiliar figurative expressions. From a cross-linguistic perspective, Deignan, Gabrys ´ , and Solska (1997) sug- gest awareness-raising activities for Polish learners of English and develop strate- gies for comprehending and creating metaphors in the second language (see also section 4.5). Another important area of applied cognitive linguistic research includes work on phrasal verbs and verb particles. In this respect, Rudzka-Ostyn (2003)workedout materials stimulating learners to develop strategies in order to grasp the meaning of English phrasal verbs and particles that are used metaphorically (see especially section 5). More general discussions on the importance of the cognitive linguistic approach to phrasal verbs and phraseological expressions can be found in Dirven (2001), Kurtyka (2001), and Queller (2001). Finally, the more general study of idiomaticity is dealt with in Ko ¨ vecses and Szabo ´ (1996), Cornell (1999), and Ko ¨ vecses (2001). 1144 martin pu ¨ tz 4. Conceptual Frameworks and Acquisition Having outlined the relevance of some major studies for Applied Cognitive Lin- guistics, I will now examine some specific tenets of Cognitive Linguistics and relate them to pedagogical implications and Second Language Acquisition research. 4.1. Learning through Insight in Motivation: Iconicity and Language Awareness According to Ungerer and Schmid (1996: 273), ‘‘The liberation from the form/ content division is probably the most important contribution that cognitive lin- guistics has made to pedagogical grammar and language teaching.’’ In other words, natural language is not just a system consisting of arbitrary signs, as assumed in the Saussurean paradigm; instead, large areas of language structure also turn out to be motivated as part of our conceptual system. The cognitive claim is that, beyond the single lexeme, language shows a strong tendency for a structural or formal cor- respondence between a symbol’s form and its meaning. In this regard, Radden (1992) refers to Haiman’s (1985) notion of ‘‘iconicity,’’ which is said to provide an excellent case against the ‘‘dogma of arbitrariness’’ and which may explain a great deal of motivation in language use. Uncovering the iconic structure of language is closely linked with the concept of consciousness raising or language awareness, defined by Rutherford and Shar- wood Smith (1985: 274) as a ‘‘deliberate attempt to draw the learner’s attention specifically to the formal properties of the target language.’’ Although the terms ‘‘consciousness raising’’ and ‘‘language awareness’’ may be used interchangeably, the latter has a wider connotation in that it refers to knowledge about language not simply in the second language learning context, but also in the framework of first language learning and teacher education (Johnson and Johnson 1998: 85). It is the latter concept of language awareness which is particularly relevant for the purposes of Applied Cognitive Linguistics oriented language pedagogy. It not only includes recognition of second language structures but also an awareness of equivalent structures in the first language, thus allowing insight into the conceptual differences of the target system and the first language or mother tongue. From a cognitive linguistic perspective, language awareness involves making the learner aware of the semantic impact of so-called symbolic units. These include not only morphosyntactic and lexical categories, but also metaphors, idioms, and formulaic phrases. The principle of language awareness and the recognition of form- meaning pairings has especially been emphasized in applied cognitive linguistic studies on figurative expressions and language teaching. In particular, students should not be geared toward random blind memorization of symbolic units, but cognitive linguistics and applied linguistics 1145 should rather be offered explanations of the systematicity and schematic nature of idiomatic language and metaphorical expressions. When linguistic expressions are paired with their underlying conceptual metaphors, they will become more transparent to the language learner; in other words, the motivation behind their idiomatic meaning will become obvious. Furthermore, it will be a central instructional principle to take into account cross-cultural differences in metaphorical themes, and in conceptual metaphors in particular (see section 4.5). Lazar (1996), for instance, refers to the figurative meaning of different colors which may vary from one language community to an- other. Idiomatic expressions like to be green and to have green fingers may evoke certain associations in some cultures which may be different from what we find in British English, where the color ‘green’ is conventionally, though not uniquely (e.g., green-eyed ‘jealous’; greenhorn ‘inexperienced’) associated with nature and innocence. Obviously, students should be made aware of these cross-cultural differ- ences by comparing the two language systems—first language and second language— in a principled way, thereby enhancing their metaphor awareness (Boers 2000b). More instances of linguo-cultural features and their underlying metaphorical sources are discussed by MacLennan (1994: 102) and Deignan, Gabrys ´ , and Solska (1997: 354). 4.2. Context- and Usage-Based Language Learning The present subsection discusses the contextual basis of Cognitive Semantics and its implications for language pedagogy. Traditionally, from the perspective of the methodology of language teaching, three major theoretical views of language and language proficiency may be distinguished (Richards and Rodgers 1986: 16). a. The structural view. It refers to language as a system of structurally related elements for the encoding of meaning; the mastery of the elements of this system (phonological, grammatical, lexical) is seen as the target of language learning. b. The functional view. It suggests that language is a means for the expression of functional meaning. The focus is on the semantic and communica- tive dimension rather than on the structural and grammatical character- istics of language. c. The interactional view. It sees language as a vehicle for the expression of interpersonal relations and the creation and maintenance of social rela- tions. Language teaching is then organized around linguistic exchanges and conversational analysis. Clearly, the functional and, to a certain extent, the interactional approach to language learning, both of which focus on the communicative function of lan- guage, seem to be most compatible with the tenets and insights of Cognitive Linguistics. The structural view is obviously the least compatible as it is in conflict 1146 martin pu ¨ tz with the cognitive linguistic tenet that semantics determines syntax. Recently, Langacker (2001b: 143) stated that ‘‘the grounding of language in discourse and social interaction is a central if not a defining notion within the functionalist tra- dition’’ and furthermore ‘‘this is no less true for cognitive linguistics.’’ Langacker (1994, 1997, 2000, 2001b) explicitly states that the study of language in its social, cultural, and discourse context is fully compatible with his claim that a concep- tual and encyclopedic view of meaning must be contextually grounded. He even goes so far as to say that ‘‘despite its mental focus, cognitive linguistics can also be described as social, cultural, and contextual linguistics’’ (Langacker 1997: 240). Langacker identifies language and culture as facets of cognition and, at the same time, recognizes the role of context and of social interaction in language use and situated discourse. In this regard, the anthropological notion of schema (see also Ungerer and Schmid’s 1996: 45–52 discussion of cognitive and cultural models)— which implies an interdependence of language, culture, and cognition—turns out to be useful as well. Consider, for instance, Sharifian’s (2001: 125) study of Ab- original English texts. By identifying major schemas in these texts, such as the travel schema, the hunting schema, and the observing schema, Sharifian is able to show that cultural knowledge and schemas may shape or influence the concep- tualizations underlying discursive structures in Aboriginal speech; in other words, schemas are cognitive structures that can be determined by cultural experiences; they are thus reflected in linguistic expressions or discourse patterns. From a language teaching perspective, then, differences in cultural practices and conceptualizations may therefore lead to miscommunication between indig- enous students and the representatives of the mainstream schooling system. The functional and interactional approaches are compatible with a learner- centered and experience-based view of second language pedagogy (Richards and Rodgers 1986: 69). This experiential view is likewise inherent in the framework of Cognitive Linguistics, where it is suggested that the world is not something merely objectively given, but that it is something construed by human perception; this construal is, in turn, guided by cultural cognition, that is, by the associations and impressions which people make as part of their personal and sociocultural expe- riences. The notion of ‘‘construal’’ and its implications for language pedagogy will be discussed in the next section. 4.3. The Learner as Conceptualizer: Construal and Linguistic Choices According to Langacker (2000: 5), the meaning of an expression consists not only of the conceptual content it evokes; equally significant is how that content is con- strued. Speakers are able to construe the same content in alternate ways, which may then result in substantially different meanings; in other words, construal refers to a speaker’s choice between various alternatives. As such, linguistic production is in cognitive linguistics and applied linguistics 1147 particular to be seen as an instance of the individual speaker’s choice or construal. These construals or linguistic alternatives may be determined by (i) the specificity or precision with which a scene is portrayed; (ii) the speaker’s perspective, which includes the Figure/Ground organization and the related aspect of viewpoint; and (iii) background assumptions which the use of a linguistic form may evoke. From a cognitive point of view, this suggests that in choosing one way of expression over the other the speaker encodes certain meanings in a specific way. The notion of ‘‘construal’’ certainly has an impact on the teaching of grammar, as appears from Achard (n.d.), in which an account is given of a construal-based approach. In particular, Achard considers two causative constructions in French that differ from each other in their word order: (1) Marie a laiss ee partir Jean. Mary has let leave John. ‘Mary let John leave’ (Achard n.d.: 8) (2) Marie a laiss ee Jean partir. Mary has let John leave. ‘Mary let John leave’ In (1), the causee (John) is represented as the initiator of the leaving event, which results in the choice of a V-V-O order; in (2), John is not presented as the source of energy initiating the process, making V-O-V the favored choice. It can be seen, then, that the two different word orders reflect the speaker’s selection of a linguistic expression more than the grammatical rules of the system per se. These different perspectives must be taken into account by the language instructor. Certainly, more research on construal-based approaches to lexical and grammatical con- structions is needed. 4.4. Frequent Uses (Prototypes) and General Meaning (Schematicity) in Language Learning The cognitive notions of ‘‘prototype’’ and ‘‘schematicity’’ can be helpful in spelling out the semantic content of a word or a grammatical construction and can thereby facilitate the language learning process. Recall that, in Langacker’s terms, a pro- totype is ‘‘a typical instance of a category (and) there are degrees of membership based on degrees of similarity’’ (1987: 371), which means that some members of a category appear to be more typical and more salient than others. When considering the category ‘furniture’, we think immediately of ‘tables’ and ‘chairs’ as best ex- amples, not of ‘mirrors’ and ‘clocks’, which suggests that membership in a category is a matter of gradience. The internal structure of categories in terms of prototyp- ical or central members and noncentral or peripheral members is likewise reflected in the semantics of grammatical categories. Taylor (1993: 211), for example, pro- poses as the prototype of ‘count noun’ a three-dimensional, concrete ‘thing’ and as 1148 martin pu ¨ tz the prototype of ‘mass noun’ an internally homogenous, divisible ‘substance’ (e.g., ‘bottle’ vs. ‘beer’). This distinction between ‘thing’ and ‘substance’ may then be transferred to the domain of time and events, which can be seen as things, and to the domain of emotional states and activities, which can bee seen as a substance. In contrast to the notion of prototype, which is a typical instance of a category, a schema is ‘‘an abstract characterization that is fully compatible with all the mem- bers of the category it defines’’ (Langacker 1987: 371); in other words, the schema embodies the commonality of its members. The importance of schematic char- acterizations for language pedagogy has been shown by a study on English com- plementation carried out by Dirven (1989b). Each of the complement clauses (e.g., to-infinitive, gerund, that-clause, etc.) has a very distinct schematic value of its own, which may be given different values and meanings in concrete contexts. Consider the following example (taken from Dirven 1989b: 116): (3) a. It’s easy to park your car here. b. Parking the car is a problem. The to-infinitive in (3a) denotes a bounded single occurrence of the event (of parking one’s car) while the predicate a problem in (3b) is a more abstract notion which requires a nonspecified, unbounded construal of the event. In structural grammar, the learner has no choice but to learn by rote the many verbs, adjectives, or nouns that govern a specific complement pattern. From the perspective of Cog- nitive Linguistics, however, it will be possible to formulate a schematic charac- terization of each complement pattern—which entails various contextual senses— thus facilitating the language learning process. Let us now return to the notion of prototype and focus more closely on its function in the construal of prepositional meanings. In Lakoff’s (1987) cognitive semantic framework, prepositions may be described as radial categories or net- works which are built around central or prototypical senses and from which var- ious senses radiate outward and are linked to the central sense by such meaning- extension processes as metaphor and metonymy. Prepositions are thus polyse- mous items which have different, yet related senses, and are described by various relations in English. Such locative relations as the up-down and front-back orientations, for instance, reflect basic experiences and are regarded as ‘‘image schemas,’’ simple and basic cognitive structures which are derived from our ev- eryday interaction with the external world (see Oakley, this volume, chapter 9). 2 Making use of work by Brugman (1981) and Lakoff (1987), Lindstromberg (1996) was one of the first to apply the findings of prototype semantics to prepositional meaning, thereby indicating pedagogical applications of prepositional semantics in the field of English Language Teaching. According to the author, current En- glish Language Teaching methods and material are not satisfactory on the grounds that only a very small set of meanings for any one preposition is presented, thus neglecting that prepositional semantics is to a large extent systematically struc- tured. With regard to the preposition on, for instance, Lindstromberg presents a set cognitive linguistics and applied linguistics 1149 . organized in the following fashion: Section 2 discusses some of the main tenets of Second Language Acquisition in the light of linguistic theories and relates them to the Cognitive Linguistics. specifically, the following four major tenets of Cognitive Linguistics were dealt with in the light of one learning problem, namely that of the English modality sys- tem: (i) the unity of linguistic. constraints on the design of teaching materials and the development of effective types of exercise (e.g., multi- text and multitask activities). A more theoretical account of the implications of cognitive